Hole-transporting host-Polymer series consisting of triphenylamine basic structures for phosphorescent polymer light-Emitting diodes

Hole-Transporting Host-Polymer Series Consisting of Triphenylamine

Basic Structures for Phosphorescent Polymer Light-Emitting Diodes

2

MANUEL W. THESEN,¹BIANCA HOFER, MARC DEBEAUX,³SILVIA JANIETZ,¹ARMIN WEDEL,¹ANNA KOHLER,

¨

HANS-HERMANN JOHANNES,³HARTMUT KRUEGER¹

¹Fraunhofer-Institut fuer Angewandte Polymerforschung, Wissenschaftspark Golm, Geiselbergstr. 69, D-14476 Potsdam, Germany

²Universitat Bayreuth, Universitatsstr. 30, 95440 Bayreuth, Germany

¨

³Labor fur Elektrooptik am Institut fur Hochfrequenztechnik, Technische Universita¨t Braunschweig, Bienroder Weg 94, 38106

¨

Braunschweig, Germany

Received 29 January 2010; accepted 10 May 2010

DOI: 10.1002/pola.24127

Published online in Wiley InterScience (www.interscience.wiley.com).

ABSTRACT: A series of novel styrene derived monomers with

triphenylamine-based units, and their polymers have been syn-

thesized and compared with the well-known structure of poly-

mer of N,N⁰-bis(3-methylphenyl)-N,N⁰-diphenylbenzidine with

respect to their hole-transporting behavior in phosphorescent

polymer light-emitting diodes (PLEDs). A vinyltriphenylamine

structure was selected as a basic unit, functionalized at the

para positions with the following side groups: diphenylamine,

3-methylphenyl-aniline, 1- and 2-naphthylamine, carbazole, and

phenothiazine. The polymers are used in PLEDs as host poly-

mers for blend systems with the following device configura-

tion: glass/indium–tin–oxide/PEDOT:PSS/polymer-blend/CsF/Ca/

Ag. In addition to the hole-transporting host polymer, the poly-

mer blend includes a phosphorescent dopant [Ir(Me-ppy)₃] and

an electron-transporting molecule (2-(4-biphenyl)-5-(4-tert-

butylphenyl)-1,3,4-oxadiazole). We demonstrate that two poly-

mers are excellent hole-transporting matrix materials for these

blend systems because of their good overall electrolumines-

cent performances and their comparatively high glass transi-

tion temperatures. For the carbazole-substituted polymer (T_g

¼

246 ^ꢀC), a luminous efficiency of 35 cd A^ꢁ1and a brightness of

6700 cd m^ꢁ2at 10 V is accessible. The phenothiazine-function-

alized polymer (T_g¼ 220 ^ꢀC) shows nearly the same outstand-

ing PLED behavior. Hence, both these polymers outperform

the well-known polymer of N,N⁰-bis(3-methylphenyl)-N,N⁰-

diphenylbenzidine, showing only a luminous efficiency of 7.9

cd A^ꢁ1and a brightness of 2500 cd m^ꢁ2(10 V).

2010 Wiley

C

V

Periodicals, Inc. J Polym Sci Part A: Polym Chem 48: 3417–

3430, 2010

KEYWORDS: charge transport; electroluminescence; fluores-

cence; host polymer; organic light-emitting diode

INTRODUCTION Polymer light-emitting devices (PLEDs) have

inspired both researchers and investors since Burroughes

et al.¹reported the first discovery of a PLED in 1990. One

remarkable advantage of polymers for light emission in dis-

plays is the ability to use simple and low-cost fabrication

methods from solution, such as printing techniques. Large-

area displays or plane lighting with light-emitting polymers

fabricated via ink jet, stamp transfer, or roll-to-roll processes

are now conceivable.^2–5The physical effect of light emission

can be divided into two different general phenomena, fluo-

rescence and phosphorescence. Statistically, excitons in or-

ganic materials are created in a ratio of 1:3 for fluorescence

to phosphorescence.⁶Thus, the light emission of a conju-

gated polymer, generated by fluorescence, is restricted by

spin statistics to an efficiency of 25%, unless spin-dependent

exciton recombination takes place.⁷These singlet emitting

materials for PLED applications are mainly conjugated poly-

mers, such as polyfluorene or poly(p-phenylenevinylene)

derivatives.^8–12With phosphorescent transition metal com-

plexes carrying organic ligands, higher quantum efficiencies

can be obtained in a light-emitting diode than with singlet

emitting materials.¹³Some transition metal complexes, in the

first instance iridium-(III) complexes, provide very short tri-

plet lifetimes of excited states due to high radiative decay

rates, so that up to 100% internal quantum efficiencies of

these phosphorescent dopants in adapted matrices are possi-

ble.¹⁴Therefore, the higher luminous efficacy of phosphores-

cent dopants and the related increased energy efficiency

compared with conjugated fluorescent polymers are the

most promising advantages using triplets for light emission

in PLEDs. Furthermore, the tuning of the phosphorescence

emission wavelength across the visible spectrum is relatively

easy by changing a ligand of the iridium complex system.^15,16

For an application of phosphorescent complexes in PLEDs,

conjugated host polymers have been investigated. However,

the interaction of fluorescence from the conjugated polymer

with phosphorescence from the triplet dopand can cause

several problems, in particular for devices that emit in the

Correspondence to: H. Krueger (E-mail: hartmut.krueger@iap.fraunhofer.de)

C

V

Journal of Polymer Science: Part A: Polymer Chemistry, Vol. 48, 3417–3430 (2010) 2010 Wiley Periodicals, Inc.

HOLE-TRANSPORTING HOST-POLYMER SERIES, THESEN ET AL.

3417

JOURNAL OF POLYMER SCIENCE: PART A: POLYMER CHEMISTRY DOI 10.1002/POLA

blue spectral range. Another possibility is the implementa-

tion of nonconjugated polymers, e.g., poly(N-vinylcarbazole)

(PVK) or polystyrene to achieve full-color phosphorescent

displays.^17,18A comparatively simple approach for the prepa-

ration of single-layer devices is the use of a polymer blend

system, whereas the active components are mixed into a

polymer single layer. Highly efficient and bright PLEDs have

been reported by blending a PVK-matrix with phosphores-

cent dopants, electron-, and additional hole-transporting

molecules.^19–22Following this idea, a phosphorescent dopant

blended together with an electron and a hole transporting

material can be considered as an emissive layer in a single-

layer device. This approach opens an easy way to investigate

new polymer charge transport materials for PLEDs. There-

fore, the main interest, beside high-efficient phosphorescent

dopants, has focused on the application of new charge-trans-

porting materials. Arrays of hole-transporting small molecule

materials have been reported, e.g., star-shaped molecules.^23–25

Charge-transporting materials have to meet a wide range of

demands, primarily electronic characteristics in respect to ef-

ficiency and brightness, adaptable energy levels, high purity,

availability, processability, high glass-transition temperatures,

and a long-term stability.

as electron-transporting material and Ir(2-(4-tolyl)pyridi-

nato)₃[Ir(Me-ppy)₃] as phosphorescent dopant. The blends

are investigated with respect to their electroluminescent

behavior in single-layer PLEDs. Finally, we report on the elec-

trochemical and optical characteristics of these polymer mate-

rials and compare the results with the well-known polymer of

N,N⁰-bis(3-methylphenyl)-N,N⁰-diphenylbenzidine (poly(TPD)).

EXPERIMENTAL

Materials and Methods

All materials were obtained from Aldrich Chemical Company

(Munich, Germany) or Acros Organics (Geel, Belgium) and used

without any further purification unless otherwise stated. Silica

gel 60 (Merck) (Darmstadt, Germany) was used in the separa-

tion and purification of compounds by column chromatography.

Solvents for column chromatography, recrystallization, and puri-

fication were received from Th. Geyer GmbH (Berlin, Germany)

and J. T. Baker (Deventer, Netherlands). Dry solvents were

received from Aldrich Chemical Company, stored over molecu-

lar sieve and sealed under an inert atmosphere.

Procedure for activation of copper: Copper (10 g, dendritic)

was stirred in 100 mL acetone with 2% of iodine for 10

min. The suspension was filtered and washed with acetone,

stirred for another 5 min in a mixture of 100 mL acetone

and 100 mL conc. HCl, filtered, and washed again with ace-

tone until neutrality. The activated copper was dried under

Polystyrene is known to be a chemically stable and optically

inert polymer. Therefore, Suzuki et al.²⁶reported a polymer

matrix system based on polystyrene with covalently fixed

electron- and hole-transporting side functionalities. Further,

to avoid phase separation and degradation during PLED

operation, the phosphorescent complex was attached to the

polystyrene backbone and for those systems external quan-

tum efficiencies up to 12% were observed. Recently, we

investigated the influence of a side-chain spacer separating

the electronically active moieties, such as charge-transporting

molecules and emitter, from the polystyrene backbone of the

terpolymers.²⁷Therefore, we built up PLEDs with these poly-

mers and compared their performances with that one of the

same polymers without these spacer moieties. It was con-

cluded that energy transfer is favored in more compact poly-

mer systems and that such transfer is suppressed when

spacer moieties are introduced in polymers with a styrene

backbone. Similarly, McKeown et al.²⁸reported on polymer

hole-transporting materials from substituted styrenes with a

triarylamine basic structure tested with respect to their

field-effect mobility in organic transistors. Furthermore,

Debeaux et al.²⁹published a new class of charge-transport-

ing polymers based on phenylbenzoimidazole moieties used

as electron-transporting material in a polymer blend system

single-layer phosphorescent PLED. The novel polymer sys-

tems showed similar electronic behavior to the small mole-

cule 1,3,5-tris(1-phenyl-1H-benzo[d]imidazol-2-yl)benzene.

ꢀ

reduced pressure at 80 C.

Thin-layer chromatography (TLC) was done with ALUGRAM

SIL G/UV₂₅₄TLC plates on aluminum from Macherey-Nagel.

Reverse-phased TLC was done with RP-18 F_254sTLC plates

on aluminum from Merck.

High-resolution (500 MHz) ¹H-NMR and ¹³C-NMR (125 MHz)

spectra were recorded on a UNITY INOVA 500 spectrometer

from Varian at room temperature.

Elemental analyses were obtained using a Thermo Scientific

FlashEA 1112 CHNS/O Automatic Elemental Analyzer.

ꢀ

Gel-permeation chromatography (GPC) at 25 C in tetrahydro-

furan (THF) has been applied to determine the molecular

weights. For this purpose, a combination of Waters HPLC-

Pump 515, Autosampler 717 plus, Dual k Absorbance Detector

2487, and a Refrative Index Detector 2414 has been used. A

precolumn and three columns from Waters (7.8 mm ꢂ 300

mm; Styragel HR3, HR4, HR5) filled with a copolymer of sty-

rene and divinylbenzene 5 lm and PS-Standards from Poly-

mer Laboratories (Varian) were used. The polymer solutions

(2 mg L^ꢁ1in THF) stirred for 24 h at room temperature and

filtered through a 1-lm syringe PTFE filter before 2 ꢂ 100 lL

of each polymer solution were injected. Molecular weights

were calculated with the Empower software from Waters.

Following this method, we established a series of novel hole-

transporting monomers by attaching these materials directly

at the polystyrene chain. In this work, we present the mono-

mer and polymer syntheses and report on the electrochemi-

cal, optical, and electroluminescent behavior of the prepared

polymers as host materials in blend systems with 2-(4-

biphenyl)-5-(4-tert-butylphenyl)-1,3,4-oxadiazole (tert-BuPBD)

Thermal analysis was performed with differential scanning

calorimetry using a Netzsch DSC 204 with a scanning rate of

10 K min^ꢁ1

.

For UV/vis and photoluminescence spectra, about 100-nm-

thick films were spun at 900 rpm from toluene solutions

3418

INTERSCIENCE.WILEY.COM/JOURNAL/JPOLA

ARTICLE

(for 8a, 8b, 8c, and 8e) or THF solutions (for 8d and 8f,

receive polymer layer thicknesses of about 70 to 90 nm after

another annealing step for 10 min at 110 ^ꢀC. The polymer-

blend solutions (62 wt % polymer, 30 wt % tert-BuPBD, 8

wt % Ir(Me-ppy)₃; the phosphorescent dopant received from

American Dye Source, Inc., were dissolved in chlorobenzene

for 24 h at room temperature using a shaker before the

polymer solutions were filtrated through polytetrafluoroeth-

ylene PTFE-syringe filters with pore sizes of 0.1 lm. The

cathodes (CsF 4 nm; Ca 15 nm; Ag 50 nm) were assembled

by a thermal evaporation process at pressures below 10^ꢁ5

Pa. Four devices of each polymer were built to guarantee

uniformity. The complete sample preparation and measure-

ments have been investigated under an inert atmosphere.

because of poor solubility in toluene) (concentrations 20 g

L

^ꢁ1) onto silica glass substrates. Absorption was measured

using a Carry 5000 UV/vis spectrometer. During the photolu-

minescence measurements, the samples were held in a con-

tinuous flow He cryostat and measured under vacuum (for

steady-state measurements) or in a small background pres-

sure of He (for temperature-dependent time-resolved mea-

surements). The temperature was recorded with an Oxford

Temperature controller ITC 502. For time-resolved measure-

ments, excitation was provided by the frequency tripled out-

put of a Nd:YAG laser at 355 nm with a pulse duration of

about 10 ns. A home-built electronic shutter was used to

reduce the excitation frequency from the 10 Hz repetition

rate of the laser to 0.2 Hz. For the steady-state spectra, the

355 and 364 nm lines of an Ar^þlaser were used as an exci-

tation source. The emission was detected with a CCD camera

Andor iDus coupled to an Oriel spectrograph MSH201 (for

steady-state spectra) or with a gated intensified ICCD camera

Andor iStar coupled to an Oriel spectrograph MS257 (for

time-resolved detection). Measurements of the photolumines-

cence quantum yields of the hosts were carried out using

the integrating sphere technique.³⁰

Device Characterization

The current–voltage and the luminance–voltage characteris-

tics of the devices were measured simultaneously with a

computer-controlled Keithley 236 Source-Measure-Unit in

combination with an Optometer Model GO352 equipped with

a calibrated sensor head for the luminance measurements.

The electroluminescence spectra were recorded with

diode-array spectrometer EPP 2000 from Stella Net Inc.

a

Field Effect Transistors

The organic field effect transistor (OFET) structures prepared

were of a ‘‘bottom gate’’-configuration using heavily doped Si-

wafers as substrates and common gate electrodes. The gate

dielectric was thermally grown directly on the wafer SiO₂of

thickness 230 nm and capacitance 14.6 nF cm^ꢁ2. On the top

of SiO₂, the source and drain electrodes (Au) were deposited

using a photolithographic technique (L ¼ 10 lm, W ¼ 195

mm). The wafers were carefully cleaned with several solvents,

dried under a nitrogen flow, and treated with hexamethyldisi-

loxane. Finally, the active polymer layers were spin-coated

from chloroform or chlorobenzene/chloroform (1:1) solutions

with polymer concentrations of 7 to 10 g L^ꢁ1under inert

atmospheric conditions. Before electrical characterization, the

samples were annealed in inert atmosphere for 5 min at 100

^ꢀC. The OFET characteristics were measured in dark under

glove box conditions, using two Source-Measure Units 236

combined with a Trigger-Control Unit 2361 and Metrics Soft-

ware (all three items Keithley Instruments).

Photoelectron spectroscopy was applied for the estimation of

the HOMO energy levels using a Riken Keiki AC-2. The inves-

tigated polymers were measured as powder stored in stain-

less steel crucibles.

IR spectra were recorded of the solids on a Digilab Scimitar

FTS2000 FT-IR spectrometer equipped with a Golden Gate

Mk II ATR-System from Specac.

Cyclovoltammetric Measurements

Voltammograms were obtained on an EG and G Parc model

273 potentiostat.

A

three-electrode configuration was

applied, contained in an undivided cell consisting of a glassy

carbon electrode (area 0.5 cm²) in which the polymer film

was deposited, a platinum mesh as the counter electrode

and an Ag/AgCl (3 M NaCl and sat. AgCl) as a reference elec-

trode. Bu₄NBF₄(0.1 M) in acetonitrile was used as an elec-

trolyte, and prior each measurement the electochemical cell

has been deoxygenated with nitrogen. The electrochemical

cell was calibrated by the use of a ferrocene standard and

the ferrocene half-wave potential has been estimated to be

435 mV for this assembly. Polymer solutions (1 wt %) in

CHCl₃were prepared and 5 lL were deposited on the glassy

carbon electrode. The prepared electrodes were kept under

Synthesis

Monomer Synthesis

4-[N,N-Di(4-iodophenyl)amino)benzaldehyde (2) [28]. Yield

96%, mp 143–144 ^ꢀC. ¹H NMR (500 MHz, CDCl₃, d, ppm):

9.84 (s, 1H, CHO), 7.71 (d, J ¼ 8.8, 2H), 7.28 (d, J ¼ 8.8, 4H),

7.06 (d, J ¼ 8.8, 2H), 6.89 (d, J ¼ 8.8, 4H). IR (ATR, cm^ꢁ1): m

¼ 3058, 3036, 3006 (w, CAH aromatic), 2794, 2715, (w,

CAH aldehyde), 1685 (m, C¼¼O), 1676 (m). Anal. calcd for

ꢀ

vacuum and dried at 60 C for 2 h.

Device Fabrication

The transparent indium–tin–oxide-covered glass slides were

received from OPTREX Europe GmbH (sheet resistance

20 X/h) and were chemically wet cleaned in an ultra sonic

bath before further processing. The hole injecting layer was

polyethylenedioxythiophene doped with polystyrenesul-

phonic acid (PEDOT:PSS,_ꢀCH8000, H.C. Starck) and dried at

temperatures above 130 C for 5 min to remove the residual

solvent. Followed by a spin-coating process the emitting poly-

mer-blend solution was deposited from chlorobenzene to

C

₁₉H₁₃I₂NO (M_w525.12 g mol^ꢁ1): C 43.46, H 2.50, N 2.67, O

3.05; found C: 43.34, H 2.49, N 2.70, O 3.17.

General Technique I: 4-{N,N-Bis[4-(N,N-diphenylamino)

phenyl]amino}benzaldehyde (3a). To simplify matters, the

reaction procedures for 3b–f were calculated in regard to

the reaction 3a. In the following, a factor is given in the

form of eq x.xx to represent the equimolar factor between

diphenylamine as amine in reaction 3a to the similar

HOLE-TRANSPORTING HOST-POLYMER SERIES, THESEN ET AL.

3419

JOURNAL OF POLYMER SCIENCE: PART A: POLYMER CHEMISTRY DOI 10.1002/POLA

reactions with the other amines for the reactions 3b–f. Com-

pound 2 5.00 g (9.52 mmol), 9.67 g (57.1 mmol) of diphenyl-

amine as amine, 10.53 g (76.2 mmol) of K₂CO₃, 3.67 g (57.1

mmol) of activated copper bronze (see above), and 0.50 g

(1.90 mmol) 18-crown-6 were weighed into a reaction flask

with a septum. The apparatus was deoxygenized by an evacu-

ation-flushing process with Ar which was repeated three times

and 50 mL of dry o-dichlorobenzene were injected via the

septum. The mixture was stirred 48 h at reflux temperature,

cooled to room temperature afterward, filtered, and concen-

trated using a rotatory evaporator. Column chromatography

with toluene led to 4.40 g (76%) of the pure product as a

slight yellow glass. ¹H NMR (500 MHz, CDCl₃, d, ppm): 9.79

(s, 1H, CHO), 7.68 (d, J ¼ 8.6, 2H), 7.27 (m, 10H), 7.13 (m,

8H), 7.04 (m, 8H), 7.02 (m, 4H). R_f¼ 0.41 (toluene). IR (ATR,

cm^ꢁ1): m ¼ 3059, 3034, 3008 (w, CAH aromatic), 2806, 2715

(w, CAH aldehyde), 1688 (m, C¼¼O). Anal. calcd for C₄₃H₃₃N₃O

(M_w607.74 g mol^ꢁ1): C 84.98, H 5.47, N 6.91, O 2.63; found:

C 84.42, H 5.47, N 6.69, O 2.98.

3059, 3034, 3008 (w, CAH aromatic), 2814, 2729 (w, CAH

aldehyde), 1686 (m, C¼¼O). Anal. calcd for C₅₁H₃₇N₃O (M_w

707.86 g mol^ꢁ1): C 86.53, H 5.27, N 5.94, O 2.26; found: C

85.56, H 5.90, N 5.21, O 2.43.

4-(N,N-Bis{4-[N,N-phenyl(2-naphthyl)amino]phenyl}amino)

benzaldehyde (3d). Using the general technique I (eq 1.00)

with 2-naphthylamine as amine and similar other reactants,

followed by column chromatography with toluene, 2.75 g

(41%) of the pure product as a slight yellow glass was

obtained. ¹H NMR (500 MHz, CDCl₃, d, ppm): 9.80 (s, 1H,

CHO), 7.75 (t, J ¼ 9.1, 4H), 7.69 (d, J ¼ 8.8, 2H), 7.62 (d, J ¼

7.8, 2H), 7.47 (s, 2H), 7.38 (m, 4H), 7.31 (m, 6H), 7.18 (d,

J ¼ 7.8, 4H), 7.10 (m, 8H), 7.06 (m, 4H). ¹³C NMR (125 MHz,

CDCl₃, d, ppm): 190.30 (1C, CHO), 153.45 (1C, ArC_q), 147.48

(2C, ArC_q), 145.10 (2C, ArC_q), 144.94 (2C, ArC_q), 140.25 (2C,

ArC_q), 143.36 (2C, ArC_q), 131.39 (2C, ArC_q), 130.15 (2C, ArCH),

129.40 (4C, ArCH), 129.04 (2C, ArCH), 128.46 (1C, ArC_q),

127.58 (2C, ArCH), 127.21 (4C, ArCH), 126.93 (2C, ArCH),

126.35 (2C, ArCH), 124.66 (2C, ArCH), 124.64 (4C, ArCH),

124.57 (4C, ArCH), 124.46 (2C, ArCH), 123.27 (2C, ArCH),

120.64 (2C, ArCH), 118.22 (2C, ArCH). R_f¼ 0.46 (toluene).

IR (ATR, cm^ꢁ1): m ¼ 3059, 3034, 3008 (w, CAH aromatic),

2806, 2723 (w, CAH aldehyde), 1690 (m, C¼¼O). Anal. calcd

for C₅₁H₃₇N₃O (M_w707.86 g mol^ꢁ1): C 86.53, H 5.27, N

5.94; found: C 85.23, H 5.27, N 5.88.

4-(N,N-Bis{4-[N,N-phenyl(m-tolyl)amino]phenyl}amino)

benzaldehyde (3b). Using the general technique I (eq

1.00) with 3-methyl-N-phenylaniline as amine and similar

other reactants, followed by column chromatography with

toluene, 2.74 g (45%) of the pure product as a slight yellow

glass were obtained. ¹H NMR (500 MHz, CDCl₃, d, ppm):

9.78 (s, 1H, CHO), 7.67 (d, J ¼ 8.6, 2H), 7.25 (m, 4H), 7.16 (t,

J ¼ 7.8, 2H), 7.11 (d, J ¼ 8.6, 4H), 7.04 (m, 12H), 6.92 (m,

4H), 6.68 (d, J ¼ 8.6, 2H), 2.28 (s, 3H). ¹³C NMR (125 MHz,

CDCl₃, d, ppm): 190.05 (1C, CHO), 153.32 (1C, ArC_q), 147.36

(2C, ArC_q), 147.16 (2C, ArC_q), 144.91 (2C, ArC_q), 139.55 (2C,

ArC_q), 138.98 (2C, ArC_q), 131.11 (2C, ArCH), 129.02 (4C,

ArCH), 128.89 (2C, ArCH), 128.01 (1C, ArC_q), 126.92

(4C, ArCH), 125.00 (2C, ArCH), 124.05 (4C, ArCH), 123.99

(4C, ArCH), 123.85 (2C, ArCH), 122.62 (2C, ArCH), 121.56

(2C, ArCH), 117.70 (2C, ArCH). R_f¼ 0.45 (toluene). IR (ATR,

cm^ꢁ1): m ¼ 3059, 3034, 3008 (w, CAH aromatic), 2920,

2853 (w), 2808, 2729 (w, CAH aldehyde), 1688 (m, C¼¼O).

Anal. calcd for C₄₅H₃₇N₃O (M_w, 635.79 g mol^ꢁ1): C 85.01, H

5.87, N 6.61, O 2.52; found: C 84.24, H 5.99, N 6.31, O 3.12.

N,N-Bis[4-(carbazolyl)phenyl]benzaldehyde (3e). Using

the general technique I (eq 2.40) with carbazole as amine

(double amount) and similar other reactants, followed by

column chromatography with toluene, 8.16 g (59%) of the

pure product as a slight yellow glass was obtained. ¹H NMR

(500 MHz, CDCl₃, d, ppm): 9.90 (s, 1H, CHO), 8.16 (d, J ¼ 7.8

Hz, 4H), 7.83 (d, J ¼ 8.6 Hz, 2H), 7.60 (d, J ¼ 8.6 Hz, 4H),

7.52–7.42 (m, 12H), 7.33–7.27 (m, 6H). R_f¼ 0.50 (toluene).

IR (ATR, cm^ꢁ1): m ¼ 3045, 3021, 3008 (w, CAH aromatic),

2712, 2688 (w, CAH aldehyde), 1687 (m, C¼¼O). Anal. calcd

for C₄₃H₂₉N₃O (M_w603.71 g mol^ꢁ1): C 85.55, H 4.84, N 6.96,

O 2.65; found: C 84.76, H 4.70, N 7.07, O 2.87.

4-{Bis[4-(phenothiazinyl)phenyl]amino}benzaldehyde

(3f). Using the general technique I (eq 0.80) with pheno-

thiazine as amine and similar other reactants, followed by

column chromatography with toluene, 3.00 g (59%) of the

pure product as a slight yellow glass was obtained. ¹H NMR

(500 MHz, CD₂Cl₂, d, ppm): 9.89 (s, 1H, CHO), 7.81 (d, J ¼

8.6, 2H), 7.38 (m, 8H), 7.24 (m, 2H), 7.08 (m, 4H), 6.96 (m,

4H), 6.88 (m, 4H), 6.42 (d, J ¼ 8.3, 4H). R_f¼ 0.50 (toluene).

IR (ATR, cm^ꢁ1): m ¼ 3061, 3034, 3008 (w, CAH aromatic),

2828, 2729 (w, CAH aldehyde), 2691, 1734 (w), 1688 (m,

C¼¼O). Anal. calcd for C₃₄H₂₉N₃OS₂(M_w667.84 g mol^ꢁ1): C

77.33, H 4.38, N 6.29; found: C 78.30, H 4.64, N 5.69.

4-(N,N-Bis{4-[N,N-phenyl(1-naphthyl)amino]phenyl}amino)

benzaldehyde (3c). Using the general technique

I (eq

0.28) with 1-naphthylamine as amine and similar other reac-

tants, followed by column chromatography with toluene,

1.11 g (59%) of the pure product as slight a yellow glass

was obtained. ¹H NMR (500 MHz, CDCl₃, d, ppm): 9.75 (s,

1H, CHO), 7.95 (d, J ¼ 8.6, 2H), 7.89 (m, J ¼ 8.3, 2H), 7.78

(d, J ¼ 8.3, 2H), 7.63 (m, J ¼ 8.8, 2H), 7.48 (m, 4H), 7.37 (m,

4H), 7.20 (m, 4H), 7.05 (m, 4H), 6.98 (m, 8H), 6.94 (m, 4H).

¹³C NMR (125 MHz, CDCl₃, d, ppm): 190.28 (1C, CHO),

153.64 (1C, ArC_q), 148.03 (2C, ArC_q), 145.74 (2C, ArC_q),

143.17 (2C, ArC_q), 139.18 (2C, ArC_q), 135.29 (2C, ArC_q),

131.34 (2C, ArC_q), 131.21 (2C, ArCH), 129.16 (4C, ArCH),

128.46 (2C, ArCH), 128.06 (1C, ArC_q), 127.31 (2C, ArCH),

127.23 (4C, ArCH), 126.70 (2C, ArCH), 126.47 (2C, ArCH),

126.36 (2C, ArCH), 126.21 (2C, ArCH), 124.10 (2C, ArCH),

122.44 (4C, ArCH), 121.88 (2C, ArCH), 121.81 (4C, ArCH),

117.62 (2C, ArCH). R_f¼ 0.41 (toluene). IR (ATR, cm^ꢁ1): m ¼

4-{N,N-Bis[4-(N,N-diphenylamino)phenyl]amino}vinylphenyl

(4a). A total of 1.24 g (3.46 mmol) MePPh₃Br was weighed

into a reaction flask with septum and the apparatus was

deoxygenized by an evacuation-flushing process with Ar

which was repeated three times. Fifteen milliliters of dry tol-

ꢀ

uene was added and the solution was cooled to 0 C. A total

of 2.16 mL (3.46 mmol) of n-BuLi (1.6 M in hexane) was

3420

INTERSCIENCE.WILEY.COM/JOURNAL/JPOLA

ARTICLE

injected slowly and the reaction mixture stirred for 20 min

at 0 C. Two grams (3.29 mmol) of the benzaldehyde deriva-

tive 3a was dissolved in 20 mL of dry toluene and added

with the benzaldehyde derivative 3c, followed by a recrystal-

ꢀ

lization from CH₂Cl₂/EtOH 1:2, 0.54 g (77%) of the pure

1

ꢀ

product was obtained as beige crystals (mp 220–223 C). H

NMR (500 MHz, THF-d₈, d, ppm): 7.95 (d, J ¼ 8.3, 2H), 7.87

(d, J ¼ 8.1, 2H), 7.76 (d, J ¼ 8.3, 2H), 7.44 (t, J ¼ 7.8, 2H),

7.41 (t, J ¼ 7.8, 2H), 7.32 (m, 4H), 7.23 (d, 2H), 7.11 (m,

4H), 6.93 (m, 14H), 6.82 (t, J ¼ 7.3, 2H), 6.60 (dd, J ¼ 10.7,

17.6, 1H, ArCH¼¼CH₂), 5.57 (d, J ¼ 17.6, 1H, CH₂¼¼CHAr

trans), 5.03 (d, J ¼ 10.7, 1H, CH₂¼¼CHAr cis). ¹³C NMR (125

MHz, CDCl₃, d, ppm): 149.52 (2C, ArC_q), 148.59 (1C, ArC_q),

144.74 (2C, ArC_q), 144.43 (2C, ArC_q), 142.75 (2C, ArC_q), 137.25

(1C, vinyl CH), 136.39 (2C, ArC_q), 132.16 (2C, ArC_q), 131.92

(1C, ArC_q), 129.66 (4C, ArCH), 129.10 (2C, ArCH), 127.74 (2C,

ArCH), 127.65 (2C, ArCH), 127.06 (2C, ArCH), 127.00 (2C,

ArCH), 126.89 (2C, ArCH), 126.75 (2C, ArCH), 126.10

(4C, ArCH), 124.90 (2C, ArCH), 124.15 (4C, ArCH), 122.72 (2C,

ArCH), 121.79 (2C, ArCH), 121.66 (4C, ArCH),111.34 (1C, vinyl

CH₂). R_f¼ 0.47 (hexane/ethyl acetate 10:1), 0.51 (THF/MeOH/

H₂O 8:8:1). IR (ATR, cm^ꢁ1): m ¼ 3086 (w, CAH vinyl), 3058,

3040, 3004 (w, CAH aromatic), 2977 (w, CAH vinyl), 1627

(w, C¼¼C vinyl). Anal. calcd for C₅₂H₃₉N₃(M_w705.89 g mol^ꢁ1):

C 88.48, H 5.57, N 5.95; found: C 87.03, H 5.81, N 5.64.

slowly to the reaction mixture. The mixture was stirred for

another 20 min at 0 C and over night at room temperature

ꢀ

afterward. The mixture was diluted with CH₂Cl₂, washed sev-

eral times with water, dried over Na₂SO₄, filtrated, and con-

centrated. Recrystallization from toluene/EtOH 2:1 gave 0.64

g (32%) of the pure product as beige crystals (mp 197–199

^ꢀC). ¹H NMR (500 MHz, THF-d₈, d, ppm): 7.28 (d, J ¼ 8.6,

2H), 7.20 (m, 8H), 7.05 (d, 8H), 6.99 (m, 8H), 6.94 (m, 6H),

6.63 (dd, J ¼ 10.7, 17.6, 1H, CH₂¼¼CHAr), 5.60 (d, J ¼ 17.6,

1H, CH₂¼¼CHAr trans), 5.06 (d, J ¼ 10.7, 1H, CH₂¼¼CHAr cis).

R_f¼ 0.59 (hexane/ethyl acetate 10:1), 0.61 (THF/MeOH/

H₂O 8:8:1). IR (ATR, cm^ꢁ1): m ¼ 3085 (w, CAH vinyl), 3061,

3034, 3008 (w, CAH aromatic), 2979 (w, CAH vinyl), 1624

(w, C¼¼C vinyl). Anal. calcd for C₄₄H₃₅N₃(M_w605.77 g

mol^ꢁ1): C 87.24, H 5.82, N 6.94; found: C 87.25, H 6.09, N

6.85.

General Technique II: 4-(N,N-Bis{4-[N,N-phenyl(m-tolyl)

amino]phenyl}amino)vinylphenyl (4b). To simplify matters,

the reaction procedures for 4c–f and 7 were calculated in

regard to the reaction 4b. In the following, a factor is given

in the form of (eq x.xx) to represent the equimolar factor

between the reaction 4b with 3b as aldehyde to the similar

reactions with the other aldehydes for the reactions 4c–f

and 7. A total of 0.58 g (5.14 mmol) of KO^tBu and 1.70 g

(4.77 mmol) of MePPh₃Br were weighed into a reaction flask

with a septum. The apparatus was deoxygenized by an evac-

uation-flushing process with Ar for three times. Twenty

milliliters of dry THF was added and the solution was cooled

to 0 ^ꢀC and stirred for 20 min. Two grams (3.15 mmol) of

the benzaldehyde derivative 3b were dissolved in 10 mL of

dry THF and injected slowly. The reaction mixture was

4-(N,N-Bis{4-[N,N-phenyl(2-naphthyl)amino]phenyl}amino)

vinylphenyl (4d). Using the general technique II (eq 0.63)

with the benzaldehyde derivative 3d, followed by a recrystal-

lization from CH₂Cl₂/EtOH 1:2, 1.21 g (86%) of the pure

product was obtained as beige crystals (mp 268–270 ^ꢀC).

¹H NMR (500 MHz, THF-d₈, d, ppm): 7.71 (m, 4H), 7.58 (d, J

¼ 7.8, 2H), 7.42 (s, 2H), 7.28 (m, 12H), 7.11 (d, 4H), 7.04

(m, 10H), 6.98 (d, 2H), 6.64 (dd, J ¼ 10.7, 17.6, 1H,

ArCH¼¼CH₂), 5.62 (d, J ¼ 17.6, 1H, CH₂¼¼CHAr trans), 5.07

(d, J ¼ 10.7, 1H, CH₂¼¼CHAr cis). ¹³C-NMR (125 MHz, CDCl₃,

d, ppm): 148.78 (2C, ArC_q), 148.41 (1C, ArC_q), 146.36 (2C,

ArC_q), 143.87 (2C, ArC_q), 143.70 (2C, ArC_q), 137.23 (1C, vinyl

CH), 135.55 (2C, ArC_q), 132.49 (1C, ArC_q), 131.04 (2C, ArC_q),

129.97 (4C, ArCH), 129.57 (2C, ArCH), 128.16 (2C, ArCH),

127.80 (2C, ArCH), 127.58 (2C, ArCH), 126.85 (2C, ArCH),

126.20 (4C, ArCH), 126.04 (4C, ArCH), 125.02 (2C, ArCH),

124.75 (6C, ArCH), 123.45 (4C, ArCH), 120.57 (2C, ArCH),

111.70 (1C, vinyl CH₂). R_f¼ 0.57 (hexane/ethyl acetate

10:1), 0.59 (THF/MeOH/H₂O 8:8:1). IR (ATR, cm^ꢁ1): m ¼

3085 (w, CAH vinyl), 3058, 3036, 3004 (w, CAH aromatic),

2985 (w, CAH vinyl), 1627 (w, C¼¼C vinyl). Anal. calcd for

ꢀ

stirred at 0 C for 1 h. CH₂Cl₂were added, the solution was

washed with H₂O, dried over Na₂SO₄, filtrated, and concen-

trated. Recrystallization from CH₂Cl₂/EtOH 1:2 led to 1.42 g

(71%) of the pure product as a beige solid (mp 185–189

^ꢀC). ¹H NMR (500 MHz, THF-d₈, d, ppm): 7.28 (d, J ¼ 8.6,

2H), 7.19 (m, 4H), 7.11 (t, J ¼ 7.8, 2H), 7.01 (m, 12H), 6.93

(m, 6H), 6.84 (d, J ¼ 7.8, 2H), 6.79 (d, J ¼ 7.3, 2H), 6.63 (dd,

J ¼ 10.7, 17.6, 1H, ArCH¼¼CH₂), 5.60 (d, J ¼ 17.6, 1H,

CH₂¼¼CHAr trans), 5.06 (d, J ¼ 10.7, 1H, CH₂¼¼CHAr cis). ¹³C-

NMR (¹²⁵MHz, CDCl₃, d, ppm): 148.90 (2C, ArC_q), 148.68 (2C,

ArC_q), 148.48 (1C, ArC_q), 144.04 (2C, ArC_q), 143.30 (2C, ArC_q),

139.66 (2C, ArC_q), 137.23 (1C, vinyl CH), 132.26 (1C, ArC_q),

129.80 (4C, ArCH), 129.72 (2C, ArCH), 127.74 (2C, ArCH),

125.98 (4C, ArCH), 125.86 (4C, ArCH), 125.42 (2C, ArCH),

124.34 (4C, ArCH), 124.24 (2C, ArCH), 123.22 (2C, ArCH),

122.96 (2C, ArCH), 122.06 (2C, ArCH), 111.60 (1C, vinyl CH₂).

R_f¼ 0.58 (hexane/ethyl acetate 10:1), 0.53 (THF/MeOH/H₂O

8:8:1). IR (ATR, cm^ꢁ1): m ¼ 3084 (w, CAH vinyl), 3061, 3034,

3008 (w, CAH aromatic), 2976 (w, CAH vinyl), 1627 (w, C¼¼C

vinyl). Anal. calcd for C₄₆H₃₉N₃(M_w633.82 g mol^ꢁ1): C 87.17,

H 6.20, N 6.63; found: C 86.79, H 6.34, N 6.40.

C

₅₂H₃₉N₃(M_w705.89 g mol^ꢁ1): C 88.48, H 5.57, N 5.95;

found: C 87.89, H 5.52, N 5.92.

N,N-Bis[4-(carbazolyl)phenyl]-4-vinylaniline (4e). Using

the general technique II (eq 2.47) with the benzaldehyde de-

rivative 3e, followed by a recrystallization from toluene/

EtOH 1:1, 2.80 g (60%) of the pure product was obtained as

colorless crystals (mp 169–171 ^ꢀC). ¹H NMR (500 MHz,

DMSO-d₆, d, ppm): 8.25 (d, J ¼ 7.7 Hz, 4H), 7.60 (d, J ¼ 8.7

Hz, 4H), 7.54 (d, J ¼ 8.6 Hz, 2H), 7.48–7.43 (m, 8H), 7.40

(d, J ¼ 8.7 Hz, 4H), 7.32–7.23 (m, 6H), 6.75 (dd, J ¼ 17.6,

11.0 Hz, 1H, ArCH¼¼CH₂), 5.79 (d, J ¼ 17.6 Hz, 1H,

CH₂¼¼CHAr trans), 5.24 (d, J ¼ 11.0 Hz, 1H, CH₂¼¼CHAr cis).

R_f¼ 0.63 (hexane/ethyl acetate 10:1), 0.43 (THF/MeOH/

H₂O 8:8:1). IR (ATR, cm^ꢁ1): m ¼ 3085 (w, CAH vinyl), 3062,

4-(N,N-Bis{4-[N,N-phenyl(1-naphthyl)amino]phenyl}amino)

vinylphenyl (4c). Using the general technique II (eq 0.31)

HOLE-TRANSPORTING HOST-POLYMER SERIES, THESEN ET AL.

3421

JOURNAL OF POLYMER SCIENCE: PART A: POLYMER CHEMISTRY DOI 10.1002/POLA

3044, 3006 (w, CAH aromatic), 1625 (w, C¼¼C vinyl). Anal.

calcd for C₄₄H₃₁N₃(M_w601.74 g mol^ꢁ1): C 87.82, H 5.19, N

6.98; found: C 87.61, H 5.29, N 6.95.

CH₂¼¼CHAr trans, J ¼ 17.6), 5.16 (d, 1H, CH₂¼¼CHAr cis, J ¼

10.7), 2.35 (s, 3H); 7b (75%): 7.44 (d, 4H, J ¼ 8.3), 7.39 (d,

1H, J ¼ 8.3), 7.25 (m, 4H), 7.16 (m, 2H), 7.11 (m, 6H), 7.02

(m, 2H), 6.92 (m, 4H), 6.85 (m, 2H), 6.91 (dd, 1H,

CH₂¼¼CHAr, J ¼ 10.7, 17.6), 5.58 (d, 1H, CH₂¼¼CHAr trans, J

¼ 17.6), 5.21 (d, 1H, CH₂¼¼CHAr cis, J ¼ 10.7), 2.27 (s, 3H).

IR (ATR, cm^ꢁ1): m ¼ 3084 (w, CAH vinyl), 3061, 3031 (w,

CAH aromatic), 2970 (w, CAH vinyl). Anal. calcd for

N,N-Bis[4-(phenothiazinyl)phenyl]-4-vinylaniline (4f). Using

the general technique II (eq 0.71) with the benzaldehyde de-

rivative 3f, followed by a recrystallization from toluene/

EtOH 1:1, 0.92 g (62%) of the pure product were obtained

as colorless crystals (mp 145–148 ^ꢀC). ¹H NMR (500 MHz,

THF-d₈, d, ppm): 7.45 (d, 2H), 7.39 (d, 4H), 7.22 (d, 4H),

7.25 (d, 2H), 6.97 (d, 4H), 6.87 (t, 4H), 6.78 (t, 4H), 6.72

(dd, J ¼ 10.7, 17.6, 1H, ArCH¼¼CH₂), 6.37 (d, 4H), 5.72 (d, J

¼ 17.6, 1H, CH₂¼¼CHAr trans), 5.17 (d, J ¼ 10.7, 1H,

CH₂¼¼CHAr cis). R_f¼ 0.61 (hexane/ethyl acetate 10:1), 0.57

(THF/MeOH/H₂O 8:8:1). IR (ATR, cm^ꢁ1): m ¼ 3085 (w, CAH

vinyl), 3061, 3038, 3006 (w, CAH aromatic), 2976 (w, CAH

vinyl), 1627 (w, C¼¼C vinyl). Anal. calcd for C₄₄H₃₁N₃S₂(M_w

665.87 g mol^ꢁ1): C 79.37, H 4.69, N 6.31, S 9.63; found: C

79.05, H 4.71, N 6.02, S 9.40.

C

₄₀H₃₄N₂(M_w542.71 g mol^ꢁ1): C 88.52, H 6.31, N 5.16;

found: C 88.31, H 6.43, N 4.71.

Polymer Synthesis

General Technique III: Poly-[4-{N,N-bis[4-(N,N-diphenyla-

mino)phenyl]amino}vinylphenyl] (8a). The monomer con-

centration was set to be 100 g l^ꢁ1in freshly distilled THF and

2 mol % N,N-azobisisobutyronitrile was selected as an initia-

tor. The reaction was carried out in a glove box system under

ꢀ

nitrogen atmosphere at 50 C for 60 h. The resulting solution

was allowed to cool to room temperature and demonomerized

by repeated precipitation into a mixture of MeOH/Et₂O 2:1.

The residue were dissolved again in THF and filtered through

a syringe PTFE filter 0.2 lm. The solvent was removed in

vacuo and the solution precipitated again into MeOH/Et₂O

2:1. The residue was filtered by the use of PTFE filters pore

size 0.45 lm. The polymer was dried for 30 h at 80 ^ꢀC in

vacuo. Yield 76%. IR (ATR, cm^ꢁ1): m ¼ 3061, 3034, 3008 (w,

CAH aromatic), 2919 (w, CAH aliphatic). T_g¼ 168 ^ꢀC. k_max

(CHCl₃) ¼ 308 nm. Anal. calcd for [C₄₄H₃₅N₃]_n: C 87.24, H

5.82, N 6.94; found: C 86.63, H 5.85, N 7.08.

4-((4⁰-(Phenyl(m-tolyl)amino)biphenyl-4-yl)(m-tolyl)amino)

benzaldehyde (6). The apparatus was evacuated and flushed

with Ar three times. Twenty milliliters of dry DMF was

ꢀ

added and cooled to 0 C. A total of 2.17 mL (23.8 mmol) of

freshly distilled POCl₃was injected and the mixture was

stirred for 30 min at 0 ^ꢀC. Ten grams (19.4 mmol) of TPD

was dissolved in 20 mL of DMF and injected into the reac-

tion mixture. The mixture was heated to 80 ^ꢀC and stirred

for 2 h, while a change in the color from slight yellow to

brown was observed. The reaction mixture was allowed to

cool to room temperature and slowly added to 500 mL of 1

M Na₂CO₃(aq.). The green precipitate was filtered off, dis-

solved in CH₂Cl₂, washed with H₂O, dried over Na₂SO₄, fil-

tered again, and concentrated. The brown oily residue was

purified by column chromatography with toluene to obtain

4.85 g (46%) of pure 6 as a slight yellow glass. ¹H NMR

(500 MHz, CDCl₃, d, ppm): 6a (27%): 9.81 (s, 1H, CHO), 7.69

(d,2H, J ¼ 8.3), 7.53 (d, 2H, J ¼ 8.3), 7.46 (d, 2H, J ¼ 8.3),

7.26 (t, 2H, J ¼ 7.6), 7.19 (m, 9H), 7.12 (d, 4H, J ¼ 8.1), 7.03

(m, 1H), 6.96 (s, 1H), 6.93 (d, 1H, J ¼ 8.1), 6.86 (d, 1H, J ¼

7.6), 2.32 (s, 3H), 2.28 (s, 3H); 6b (73%): 10.06 (s, 1H,

CHO), 7.62 (d, 1H, J ¼ 8.5), 7.53 (d, 2H, J ¼ 8.3), 7.46 (d, 2H,

J ¼ 8.3), 7.35 (t, 2H, J ¼ 7.8), 7.26 (t, 2H, J ¼ 7.6), 7.19 (m,

4H), 7.12 (d, 4H, J ¼ 8.1), 7.03 (m, 3H), 6.96 (s, 1H), 6.93 (d,

1H, J ¼ 8.1), 6.90 (d, 1H, J ¼ 8.5), 6.86 (d, 1H, J ¼ 7.6), 6.81

(s, 1H), 2.54 (s, 3H), 2.28 (s, 3H). R_f¼ 0.45 (toluene). IR

(ATR, cm^ꢁ1): m ¼ 3060, 3031, 3007 (w, CAH aromatic),

2923, 2858 (w), 2839, 2730 (w, CAH aldehyde), 1735, 1682

(m, C¼¼O). Anal. calcd for C₃₉H₃₂N₂O (M_w544.68 g mol^ꢁ1): C

86.00, H 5.92, N 5.14; found: C 85.57, H 6.13, N 5.05.

Poly-[4-[N,N-bis{4-[N,N-phenyl(m-tolyl)amino]phenyl}amino)

vinylphenyl] (8b). Using the general technique III, yield

65%. IR (ATR, cm^ꢁ1): m ¼ 3059, 3034, 3008 (w, CAH aro-

matic), 2919 (w, CAH aliphatic). T_g¼ 141 ^ꢀC. k_max(CHCl₃)

¼ 316 nm. Anal. calcd for [C₄₆H₃₉N₃]_n: C 87.17, H 6.20, N

6.63; found: C 86.78, H 6.17, N 6.67.

Poly-[4-{N,N-bis[4-(N,N-phenyl(1-naphthyl)amino]phenyl}

amino]vinylphenyl] (8c). Using the general technique III,

yield 68%. IR (ATR, cm^ꢁ1): m ¼ 3058, 3036, 3007 (w, CAH

ꢀ

aromatic), 2925, 2917 (w, CAH aliphatic). T_g¼ 173 C. k_max

(CHCl₃) ¼ 331 nm. Anal. calcd for [C₅₂H₃₉N₃]_n: C 88.48, H

5.57, N 5.95; found: C 87.61, H 5.58, N 6.07.

Poly-[4-{N,N-bis[4-(N,N-phenyl(2-naphthyl)amino]phenyl}

amino]vinylphenyl] (8d). Using the general technique III,

yield 75%. IR (ATR, cm^ꢁ1): m ¼ 3062, 3042, 3014 (w, CAH

aromatic), 2918 (w, CAH aliphatic). T_g¼ 158 ^ꢀC. k_max

(CHCl₃) ¼ 326 nm. Anal. calcd for [C₅₂H₃₉N₃]_n: C 88.48, H

5.57, N 5.95; found: C 87.51, H 5.53, N 5.97.

Poly-[N,N-Bis[4-(carbazolyl)phenyl]-4-vinylaniline]

(8e). Using the general technique III, yield 85%. IR (ATR,

cm^ꢁ1): m ¼ 3062, 3045, 3012 (w, CAH aromatic), 2919 (w,

CAH aliphatic). T_g¼ 246 ^ꢀC. k_max(CHCl₃) ¼ 324 nm. Anal.

calcd for [C₄₄H₃₁N₃]_n: C 87.45, H 5.90, N 6.65; found: C

87.07, H 5.22, N 7.15.

N-Phenyl-N,N⁰-di[m-tolyl-N⁰-(4-vinylphenyl)]biphenyl-4,4⁰-

diamine (7). Using the general technique II (eq 2.69) with

the benzaldehyde derivative 6, after column chromatography

with toluene, 3.80 g (83%) of 7 was obtained as a colorless

1

ꢀ

solid (mp 85–88 C). H NMR (500 MHz, CDCl₃, d, ppm): 7a

(25%): 7.44 (d, 4H, J ¼ 8.3), 7.30 (d, 1H, J ¼ 8.3), 7.25 (m,

4H), 7.16 (m, 2H), 7.11 (m, 8H), 7.02 (m, 2H), 6.92 (m, 4H),

6.65 (dd, 1H, CH₂¼¼CHAr, J ¼ 10.7, 17.6), 5.64 (d, 1H,

Poly-[N,N-Bis[4-(phenothiazinyl)phenyl]-4-vinylaniline]

(8f). Using the general technique III, yield 75%. IR (ATR,

cm^ꢁ1): m ¼ 3058, 3035, 3007 (w, CAH aromatic), 2922,

3422

INTERSCIENCE.WILEY.COM/JOURNAL/JPOLA

ARTICLE

FIGURE 1 Synthesized hole-transporting polymers 8a–f. Reagents and conditions: (i) CH₃COOH, H₂O 10:1, KI, KIO₃, 3 h, 80 ^ꢀC—2

(ii) K₂CO₃, activated Cu-Bronze, 18-crown-6, 1,2-dichlorobenzene, amine a–f (N,N-diphenylamine a, 3-Methyl-N-phenylaniline b, N-

phenyl-1-naphthylamine c, N-phenyl-2-naphthylamine d, carbazole e, phenothiazine f), 48 h, 180 ^ꢀC—3a–f. (iii) n-BuLi, MePPh₃Br,

THF, 2 h, 0 ^ꢀC; 4a, KO^tBu, MePPh₃Br, THF, 2 h, 0 ^ꢀC—4b–f. (iv) THF, AIBN, 50 ^ꢀC, 60 h—8a–f.

2917 (w, CAH aliphatic). T_g¼ 220 ^ꢀC. k_max(CHCl₃) ¼ 312

nm. Anal. calcd for [C₄₄H₃₁N₃S₂]_n: C 79.37, H 4.69, N 6.31, S

9.63; found: C 78.36, H 4.48, N 6.12, S 9.51.

yields. Monomer 4a was synthesized with n-butyl lithium as a

base for the preparation of the respective phosphonium yield.

This reagent turned out to be rather detrimental because of

the low yield (only 30%) in the last reaction step and many

undesirable side products, which are difficult to separate from

the target structure. Therefore, subsequent reactions were car-

ried out using KO^tBu for the preparation of the monomers

4b–f, followed by a yield increase to approximately 70%.

Poly-[N-phenyl-N,N⁰-Di(m-tolyl-N⁰-(4-vinylphenyl))bi-

phenyl-4,4⁰-diamine] (9). Using general technique III, yield

55%. IR (ATR, cm^ꢁ1): m ¼ 3059, 3029, 3007 (w, CAH aro-

matic), 2920 (w, CAH aliphatic). T_g¼ 171 ^ꢀC. k_max(CHCl₃)

¼ 354 nm. Anal. calcd for [C₄₀H₃₄N₂]_n: C 88.52, H 6.31, N

5.16; found: C 87.83, H 6.39, N 5.32.

The synthesis of the reference monomer, introduced by

Suzuki et al., based on a TPD structure is shown in Figure 2.

The first reaction step was a Vilsmeier formylation of TPD

with phosphorus oxychloride in N,N-dimethylformamide to

obtain the aldehyde derivative 6. Suzuki et al. published an

RESULTS AND DISCUSSION

Monomer Syntheses

We followed a synthetic route introduced by McKeown et al.²⁸

for the monomer syntheses. The reaction scheme used for all

synthesized monomers is displayed in Figure 1. The first step

was the halogenation of 4-(diphenylamino)-benzaldehyde (1)

with potassium iodine and potassium iodate in a 10:1 mixture

of acetic acid and water to obtain compound 2. Ullmann reac-

tion with the selected amine (compound a-f) followed by a

Wittig reaction with potassium tert-butylate (KO^tBu) and

methyltriphenylphosphonium bromide (MePPh₃Br) in THF

leads to the vinyl-functionalized monomers 4b–f with good

1

isomeric mixture of 28% 6a and 72% 6b. H NMR spectros-

copy confirmed a mixture of 27% 6a and 73% 6b for our

product. The final step was a Wittig reaction with KO^tBu and

MePPh₃Br in THF to obtain the isomeric mixture of the final

vinyl-functionalized monomer 7a/b in a ratio of 1:3.

Polymerizations

The polymerizations were carried out in a glove box system

under nitrogen atmosphere, whereas THF acts as solvent

while the initiator was N,N⁰-azobisisobutyronitrile. The poly-

FIGURE 2 Synthesized hole-transporting polymer 9 (mixture of two implemented isomers 7a/b). Reagents and conditions: DMF,

POCl₃, 2 h, 80 ^ꢀC—6a/b. MePPh₃Br, KO^tBu, 2h, 0 ^ꢀC—7a/b. (iii) THF, AIBN, 50 ^ꢀC, 60 h—9.

HOLE-TRANSPORTING HOST-POLYMER SERIES, THESEN ET AL.

3423

JOURNAL OF POLYMER SCIENCE: PART A: POLYMER CHEMISTRY DOI 10.1002/POLA

TABLE 1 Overview on Analytical Data of Polymers 8 a–f and 9

Polymer

8a

8b

8c

8d

8e

8f

9

Approximate yield in %

76

65

68

75

85

75

55

GPC in 10³g mol^ꢁ1

M_n

18.3

33.4

1.82

168

15.6

26.5

1.70

141

9.5

10.5

23.6

2.24

158

41.2

197.7

4.80

246

9.8

12.9

24.6

1.90

171

M_w

21.7

2.28

173

26.1

2.54

220

PDI

DSC in ^ꢀC (T_g)

ꢀ

merization temperature was 50 C for 60 h. Table 1 gives an

overview of the polymer analyses data. The approximate

yield of all polymerizations was about 70%. On one hand the

phenyl-functionalized polymers 8a, 8b, and 9 present with

polydispersity indices (PDIs) of about 1.8 reasonably narrow

molecular weight distributions for free-radical polymeriza-

tion processes. In contrast, the polymers with phenyl-naph-

thyl-functionalities and the polymer with the phenothiazine

side group show quite higher PDIs of approximately 2.4,

which can be mainly attributed to a poor monomer solubility

observed in THF, especially for the naphthylamino-functional-

ized monomers. The maximum quotient between the weight

average molar mass M_wand the number average molar mass

M_nis obtained for polymer 8e, with the carbazole units, which

has a very high PDI of 4.80. The solubility of the carbazole-

monomer in THF is quite good, so the high PDI may be due to

a different origin. We observed a bimodular distribution in the

size exclusion chromatogram of this polymer, so the compara-

tively high PDI of 4.80 may possibly be caused by the Norrish-

Trommsdorff effect or chain-transfer mechanisms themselves.

tionally, the carbazole shows an absolutely planar structure,

whereas the phenothiazine is slightly twisted at the CAS

bonds, i.e., the carbazole perform a better packaging and a

more compact polymer chain, resulting in the higher T_gcom-

pared with that one of phenothiazine.

Spectroscopic and Electrochemical Investigation

The optical behavior of the synthesized polymers was stud-

ied in spin-coated solid films on silica glass substrates and

in chloroform solutions. The UV/vis data of the films and

chloroform solutions are summarized in Table 2, along with

the 10K phosphorescence lifetimes and 0-0 peak positions

and the 300K solid-state quantum yields. Only small shifts in

a normal range between the maxima in solid film compared

with those in solution are observable for all polymers.

To identify the singlet and triplet excited states formed in

the compounds after excitation, we have performed spectro-

scopic measurements on thin films. Figure 3(a) shows the

room temperature spectra of absorption and photolumines-

cence for continuous wave excitation. We observe a differ-

ence between the emission spectra of compounds 8c and 8d

compared with the spectra of 8a, 8b, 8e, and 8f. The latter

all show two peaks centered at about 3.0 eV (ꢃ410 nm) and

2.5 eV (ꢃ500 nm) for 8a, 8b, and 8e, and at 2.6 eV (ꢃ475

nm) and 2.1 eV (ꢃ590 nm), respectively, for 8f. In contrast,

the spectra of 8c and 8d show only one broad emission

peak centered at about 2.5 eV. In absorption, the difference

is less pronounced. We notice a slightly lower energy onset

of absorption at about 3.0 eV for 8c and 8d compared with

about 3.2–3.3 eV (ꢃ375–390 nm) for the other compounds.

The photoluminescence spectra are investigated in more

detail in Figure 3(b) in which time-dependent emission spec-

tra at 10 K are shown. The spectra were taken at delay times

of 50 ns (solid line) and 500 ns (dotted line) after an excita-

tion pulse. For both delay times, the gate width of the detec-

tor was 80 ms, i.e., any long-lived signal was collected.

Although only the long-lived component was recorded with a

delay time of 500 ns, in the measurement with 50 ns delay

time possible short-lived components are added.

For the glass transition temperatures (T_g), the phenyl-substi-

tuted polymers show the lowest T_gwith 168 C for the non-

ꢀ

substituted and 141 ^ꢀC for the methylene-substituted poly-

mer, where this difference is caused by the additional methyl

groups. Furthermore, the 1- and 2-naphthyl-derived poly-

mers exhibit a difference of 15 K in the glass transition tem-

peratures on their part, mainly caused by the higher flexibil-

ity of the 2-naphthyl side groups compared with the 1-

substitution. Nevertheless, it has to be mentioned that the

relatively small molecular weights of polymers 8c and 8d

suggest that the measured T_gis not yet independent of the

molecular weight, i.e., with similar molecular weights of 8c

and 8d compared with the other polymers, a higher T_g

should be observable. However, polymer 9 is in the same

ꢀ

range with a T_gof 171 C. The carbazole-substituted 8e (246

^ꢀC) and the polymer with phenothiazine side groups 8f (220

^ꢀC) present the highest T_gs. Comparing the structures of

these two polymers 8e and 8f with that of 8a, it is obvious

that the glass transition temperature is much higher, which

is caused by the stiffening CAC and CAS bonds, respectively,

between the aromatic rings. In 8a all three amine bonds are

free in rotation, whereas in the case of the carbazole and the

phenothiazine side groups the nitrogen is absolutely fixed,

resulting overall in a stiffening of the polymer chain. Addi-

As in the steady-state room temperature spectra, we observe

a high-energy band with little structure that is centered

roughly at about 2.9 eV (ꢃ430 nm) for compounds 8a, 8b,

and 8e, and at about 2.6 eV (ꢃ475 nm) for 8c, 8d, and 8f.

3424

INTERSCIENCE.WILEY.COM/JOURNAL/JPOLA

ARTICLE

TABLE 2 Optical and Electrochemical Properties of Polymers 8a–f and 9

Polymer

8a

8b

8c

8d

8e

8f

9

Absorption maximum^a(nm)

Absorption maximum^b(nm)

Absorption maximum^b(eV)

Luminance Quantum Yield^c(%)

LUMO^d(eV)

HOMO (CV)^e(eV)

HOMO (AC-2)^f(eV)

Optical Bandgap^g(eV)

308

321

3.86

4.0

316

321

3.86

7.5

331

333

3.72

4.5

326

324

325

3.82

5.0

312

315

3.94

0.5

354

356

3.48

330

3.76

10.0

1.97

5.12

4.98

3.01

2.30

268

1.93

5.04

5.07

3.14

2.59

256

1.93

5.07

5.08

3.15

2.60

235

2.16

5.07

5.06

2.90

2.25

123

2.08

5.40

5.41

3.33

2.70

228

2.16

5.28

5.41

3.25

2.40

65

2.32

5.32

5.40

3.08

0-0 Peak of Phosphorescence (eV)

Phoshorescence Lifetime^h(ms)

a

e

f

In CHCl₃solution.

Cyclovoltammetry.

b

As solid film.

Photoelectron spectroscopy.

c

g

h

61%.

Obtained from the onset of absorption.

At 10 K.

d

LUMO-energy levels were calculated by the use of UV/vis data

This band disappears for a delay time of 500 ns and is

therefore assigned to fluorescence.

To clarify the origin of this band, we measured its decay as a

function of time (Fig. 4). The photoluminescence decays can

be fitted by a single exponential yielding lifetimes of tens and

hundreds of milliseconds (see Table 2). On the basis of the

long lifetime and the vibrational structure, we assign the well-

resolved low-energy emission at 10 K to phosphorescence.

At 10 K, there is an additional low-energy band with a well-

resolved vibrational structure. The 0–0 peaks of the low-

energy band are at about 2.6–2.7 eV for 8a, 8b, and 8e and

at about 2.2–2.4 eV for 8c, 8d, and 8f (see Table 2 for exact

values). This low-energy band is present at both delay times,

although for compounds 8c and 8d, it is difficult to discern

in the long red tail of the emission spectra taken at 50 ns.

We note the difference in vibrational structure between the

fluorescence and the phosphorescence bands. This implies a

different electronic origin for the singlet and the triplet state.

FIGURE 3 (a) Absorption (dotted line)

and steady state photoluminescence

(PL) (solid line) spectra at room tem-

perature on solid films; (b) solid film

photoluminescence (PL) spectra at 10

K, at a delay of 50 ns (solid line) and

500 ns (dotted line) after the laser

pulse. T and S denote the assignment

to triplet and singlet excited states,

respectively.

HOLE-TRANSPORTING HOST-POLYMER SERIES, THESEN ET AL.

3425

JOURNAL OF POLYMER SCIENCE: PART A: POLYMER CHEMISTRY DOI 10.1002/POLA

In a similar manner, the shorter lifetime of 8c compared

with 8d along with the lower luminescence quantum yield

also suggests a more effective nonradiative decay. It is well

known that the triplet state in naphthalene compounds loses

its energy very efficiently and nonradiatively via the high-

energy CAH stretching vibrations.³²Furthermore, naphtha-

lene has its transition dipole moment along the short axis of

the molecule. Consequently, the attachment of the naphtha-

lene moiety to the amine in 8c along the naphthalene short

axis on the 1-position allows for an efficient coupling of the

electronic transition to the CAH stretching vibration, in con-

trast to electron redistribution that is associated with the

transition in 8d. This is further confirmed by quantum chem-

ical calculations of the electron wave functions on the basis

of AM1/ZINDO (Austin Model 1/Zerner Intermediate neglect

of differential overlap).

FIGURE 4 Decay of the photoluminescence intensity with time

in thin films at 10 K.

We now compare the spectra observed for compounds 8c

and 8d with those of 8a, 8b, 8e and 8f. At 10 K, the emission

from the two naphthalene-containing polymers is similar to

that of the other compounds except for an overall bathchro-

mic shift. This suggests that conjugation continues reasonably

well through the nitrogen atoms resulting in more extended

conjugated systems for the compounds with the naphthalene

moieties in 8c and 8d compared with the compounds with

only a short ‘‘phenyl’’ moiety attached, as in 8a, 8b, 8e, and

8f. To account for the difference in the room temperature

steady-state spectra, i.e., the lack of a second emission band

in 8c and 8d, more detailed spectroscopic investigations are

required that are beyond the scope of this study.

For example, it is conceivable for the singlet state to have n–

p* or intramolecular charge-transfer character, both leading

to little vibrational structure, while the well-resolved vibra-

tional structure suggests that the triplet state might originate

from a transition with a predominant p–p* character. Since

the exchange energy for a p–p* transition tends to be larger

than that for a n–p* or intramolecular charge-transfer transi-

1

tion, an energetic arrangement of states in the order (p–p*),

¹(n–p* or CT state), ³(n–p* or CT state), ³(p–p*) is highly

possible.⁷While this might be the case in these compounds,

quantum chemical calculations are required to clarify this.

We now consider the triplet lifetimes of the compounds in

more detail. Compounds 8a, 8b, and 8e only differ in the

choice of an additional CH₃sidegroup (8b) or a stiffening

link (8e), so it is natural for the spectroscopic behavior and

the measured lifetimes to be very similar. Accordingly, the

overall luminescence quantum yields of 8a and 8e are essen-

tially the same, 4 6 1% and 5 6 1%, respectively. The quan-

tum yield of 8e is higher (7.5 6 1%), consistent with the

higher luminous efficiency of LEDs made with 8b as host

compared with 8a [Fig. 6(b)]. This points to a beneficial

effect of the CH₃sidegroup in suppressing detrimental inter-

molecular interactions. In the 10 K lifetime, this difference is

not manifested since diffusion to sites with interchain inter-

actions is usually frozen out at such low temperatures.⁷

In addition to the excited state energies, the positions of the

frontier molecular orbitals are relevant for hole-transporting

materials. Hole injection takes place from the anode into the

highest occupied molecular orbital (HOMO). The HOMO

energy levels of the polymers were investigated by photo-

emission spectroscopy of the polymer powder and verified

by cyclovoltammetric investigations as solid films on a glassy

carbon electrode. Overall, the HOMO energy levels of the

polymers obtained with both methods are in a good agree-

ment. We derived the optical bandgap from the solid film

absorption spectra and calculated the LUMO by the use of

the HOMO measured via photoelectron spectroscopy. LUMO,

HOMO, and bandgap values are listed in Table 2 and dis-

played in Figure 5.

In compound 8f, the addition of the sulfur atom results in

an overall bathochromic shift of the excited states and a

strongly reduced lifetime for the triplet state. The lifetime of

8f is only half the lifetime of 8c, even though the triplet state

in 8c is at lower energy. Clearly, the short lifetime of 65 ms

cannot be attributed merely to the slightly lower triplet

energy according to the energy gap law.³¹Rather it seems

that the introduction of the sulfur atom introduces some

additional nonradiative or radiative decay channels. The

emission from 8f is weak and borders on the detection limit

of our quantum efficiency measurement setup. From this, we

infer that the decay route introduced by the incorporation of

the sulfur atom to the chemical structure is of an undesired

nonradiative type.

The structurally very similar phenylamino-substituted poly-

mers 8a and 8b have nearly identical oxidation and reduc-

tion potentials. Comparison of 8a with 8c shows the modifi-

cation from a phenyl group to a naphthalene group leaving

the HOMO almost unchanged. Comparing the mesomeric

structures of the two polymers 8c and 8d with the naphtha-

lene groups, 8c with the naphthalene in 1-position exhibits

more mesomeric structures retaining full aromaticity than

structure 8d with the naphthalene connected in 2-position.

As a result, the larger conjugated system 8c has a somewhat

reduced LUMO and bandgap.

When the two phenyl rings are stiffened by a C–C link to

form a carbazole as in 8e, the conjugation in this planar ring

3426

INTERSCIENCE.WILEY.COM/JOURNAL/JPOLA

ARTICLE

Device Characterization

Single-layer phosphorescent PLEDs were prepared with

these hole-transporting polymers in configuration of

a

(glass/indium–tin–oxide/PEDOT:PSS/polymer-blend/CsF/Ca/

Ag) to compare the influence of different host-polymers on

the electroluminescence properties of the device. In addition

to the hole-transporting host-polymers, the blends consisted

of 30 wt % tert-BuPBD as the electron-transporting material

and of 8 wt % Ir(Me-ppy)₃as the phosphorescent dopant.

The electroluminescent behavior of the green emitting phos-

phorescent PLEDs are summarized in Table 3 and displayed

in Figure 6. The electroluminescence spectra of the single-

layer devices are illustrated in Figure 7 and the maxima of

emission are further noted in Table 3.

FIGURE 5 Measured HOMO, calculated LUMO energy levels,

and calculated optical bandgaps of polymers 8a–f and 9.

system is enhanced compared with 8a. It was found that

two clearly separated energy levels are present at ꢁ5.41 and

ꢁ5.74 eV measured by photoelectron spectroscopy. This

result was further verified by cyclovoltammetric and the

‘‘higher’’ level is used as potential HOMO energy level in the

following discussion. For polymer 8e, the first oxidation

potential is not very distinct in the measured photoelectron

spectrum. The insertion of a sulfur atom in 8f leads to a

hypsochromic shift in the absorption maximum and the sul-

fur atom is responsible for a higher electron density in the

attached molecule. Furthermore, the phenothiazine has a dis-

turbed planar structure, the molecule is slightly angulated at

the sulfur atom, whereas, the carbazole unit is strictly planar.

This angle decreases the conjugation in the phenothiazine

moiety. Comparison with literature shows a very good ac-

cordance between the measured energy levels of these poly-

mers to those of small molecules with structural similarity,

for example (8b³³, 8c³⁴, 8d³⁵, 8d³⁶, and 9³⁷).

In the electroluminescence spectrum, devices made with the

host 8a, 8b, 8e, and 8f show essentially the same emission

than the device made with the TPD host 9, except that the

shoulder at 550 nm has a little more intensity for 8a and 8e.

Because of the energetic position and vibrational structure, we

attribute this emission band to the expected phosphorescence

from the Ir(Me-ppy)₃guest. For the hosts 8c and 8d, the

onset of emission still coincides with the Ir(Me-ppy)₃guest

emission; however, the emission peaks are shifted batho-

chromically, the spectra are broadened, and they show a tail

extending to longer wavelengths. This observation points to a

contribution from a state at lower energy such as an exciplex

or electroplex. The latter denotes an exciplex state that is only

formed for electrical excitation. To distinguish between exci-

plex and electroplex formation, we compare in Figure 7(c) the

photoluminescence from a blend of the hosts 8c with Ir(Me-

ppy)₃and 8d with Ir(Me-ppy)₃with the electroluminescence

of corresponding devices. The red emission feature is not

present for optical excitation of the blend, it only appears for

electrical excitation. The long red tail observed in devices

made with 8c and 8d can therefore be attributed to electro-

plex formation. Such electroplex emission is usually inefficient

due to the low oscillator strength of such a charge-transfer-

type state, and this may account for the poor performance of

the LEDs made from 8c and 8d with respect to luminous effi-

ciency or brightness.³²

In conclusion, the measured HOMO energy levels of these

investigated host polymer materials should allow the devel-

opment of efficient phosphorescent PLEDs. Therefore, we

decided to test PLEDs with our polymers as hole-transport-

ing host materials, Ir(Me-ppy)₃(HOMO ꢁ5.1 eV and LUMO

ꢁ2.6 eV)³⁸as phosphorescent dopant and tert-BuPBD as

electron-transporting material (HOMO ꢁ5.9 eV and LUMO

ꢁ2.4 eV).³⁹In case of 8e, 8f, and 9, with the lower HOMO,

compared with the other polymers, a more efficient hole

transfer to the emitter (HOMO ꢁ5,1 eV) should be possible.

In addition to this, we note that in our spectroscopic analysis

(Fig. 3), we find the 0–0 peak of the 10 K phosphorescence

TABLE 3 Electroluminescent Characteristics of Polymer-Blend PLEDs of 8a–f and 9

Polymer

8a

8b

8c

8d

8e

8f

9

Layer thickness (nm)

86

96

83

93

74

87

Luminous efficiency^a(cd A^ꢁ1

)

18.5

21

12.3

16

0.5

29

0.5

22

32.8

12

29.0

12

7.9

21

Current density^a(mA cm^ꢁ2

Luminous efficiency^b(cd A^ꢁ1

Brightness^c(cd m^ꢁ2

)

9.4

16.7

4200

519

0.5

200

545

0.5

150

539

35.1

6700

522

28.2

6100

520

8.7

2600

515

)

5000

517

Electroluminescence maximum (nm)

a

c

At 8 V.

At 100 cd m^ꢁ2

At 10 V.

b

.

HOLE-TRANSPORTING HOST-POLYMER SERIES, THESEN ET AL.

3427

JOURNAL OF POLYMER SCIENCE: PART A: POLYMER CHEMISTRY DOI 10.1002/POLA

FIGURE 6 (a) Brightness and current

density as function of the applied bias

of polymers 8a–f and 9 as blends with

tert-BuPBD and Ir(Me-ppy)₃; (b) Lumi-

nous efficiency as a function of the cur-

rent density of polymers 8a–f and 9 as

blends with tert-BuPBD and Ir(Me-ppy)₃

(polymer layer thicknesses noted in

Table 3).

of the hosts 8c and 8d to be at 2.25 eV (551 nm) and 2.30

eV (539 nm), respectively. In Figure 7, we find the 300 K 0–

0 peak of our phosphorescent emitter, Ir(Me-ppy)₃, to be at

2.39 eV (517 nm) with 8a or 2.40 eV (519 nm) with 8b.

This is consistent with literature values of 2.42 eV (512 nm),

and at higher energies than in the hosts 8c and 8d.³⁸An

energetic arrangement where the host triplet is at lower

energy than the guest triplet is unsuitable since any triplets

that still are formed on the guest may be quenched by the

host. This process can further contribute to poor EL

performances.

In summary, the polymer 8e with the carbazolyl side groups

exhibited the best overall performance with a luminous effi-

ciency of about 35.1 cd A^ꢁ1at 100 cd m^ꢁ2. Furthermore, 8e

has a quite low current density of 0.02 A cm^ꢁ2at the maxi-

mum voltage of 10 V. Only the phenothiazine-functionalized

polymer showed nearly similar charge transport behavior.

The electroluminescence maxima for these two polymers as

matrix material in this blend system are at 522 and 520 nm,

respectively, and reflect the phosphorescent emission of the

dopant. Because of their excellent charge transport behavior,

these two polymers have been chosen for further investiga-

tions. The best device performance for 8e and 8f is in

good relation to their measured HOMO values that allow an

efficient hole transfer to the emitter described above,

already.

We now compare the host polymers in the first group that is

8a, 8b, 8e, and 8f. We note that the luminous efficiencies at

100 cd m^ꢁ2scale with the HOMO level of the host polymers.

Highest efficiency is reached with 8e that has a HOMO of

ꢁ5.41 eV, followed by 8f (ꢁ5.41 eV), 8b (ꢁ5.08 eV), and 8a

(ꢁ5.07eV). We attribute the noticeably better performance

of 8b compared with 8a to the additional CH₃sidegroup

that can prevent possible detrimental interchain interactions.

The TPD derivative 9, which is known to be an efficient

hole-transporting matrix molecule, only shows a luminous

efficiency at 100 cd m^ꢁ2of 8.7 cd A^ꢁ1. Furthermore, an exci-

plex formation between TPD and tert-BuPBD for small

FIGURE 7 (a) Stacked, (b) super-

posed electroluminances of poly-

mers 8a–f and 9, (c) comparison

of photoluminescence of 8c and

8d (blended with Ir(Me-ppy)₃)

with electroluminescence (poly-

mer layer-thicknesses of devices

noted in Table 3).

3428

INTERSCIENCE.WILEY.COM/JOURNAL/JPOLA

ARTICLE

molecules has been reported before.⁴⁰In our case, there is

no broadening and no bathochromic shift observable in the

emissive spectra so the maxima of emission for this device

was found at 515 nm. As mentioned before, the HOMO- and

LUMO-energy levels of the hole-transporting host-polymers

are well matched to the energy levels of the phosphorescent

dopant and the electron transporting material to obtain an

efficient device, i.e., the difference in the energy levels of the

host-polymers are not sufficient to interpret the different

luminescent behavior of the devices. In conclusion, there has

to be another reason for the difference in brightness and lu-

minous efficiency of the TPD derivative to the carbazole-sub-

stituted polymer. Therefore, we prepared organic field effect

transistors of the poly(TPD) 9 and the most promising car-

bazole-functionalized polymer, 8e, to ascertain their field

effect mobility. The field effect mobility of poly(TPD) 9 was

found to be 1 ꢂ 10^ꢁ4cm²V^ꢁ1s^ꢁ1and that one of the car-

bazole-functionalized polymer 8e turned out to be limited to

rials for PLEDs. The polymers were characterized with

respect to their EL performance in single-layer green phos-

phorescent PLEDs as host-polymers in blend systems. Singlet

and triplet excited states of the host polymers were assigned

on the basis of time- and temperature-dependent photolumi-

nescence measurements. Triplet excited state energies in the

hosts range from 2.70 to 2.25 eV, with the lower energy trip-

lets in the naphthyl derivatives. Although the naphthyl deriv-

atives are therefore not suitable as hosts for the green-emit-

ting Ir(Me-ppy)₃, the other polymers showed very narrow

emission spectra, which demonstrates; high suitability for

the creation of clear colors. Two polymers turned out to be

favored for further investigations, because of their excellent

host behavior in such blend systems. The carbazole-substi-

tuted polymer, 8e, is the most efficient and brightest hole-

transporting polymer matrix in a blend system with tert-

BuPBD and Ir(Me-ppy)₃, followed by the phenothiazine com-

pound 8f. This is attributed to a good balance of electron

and hole mobility. Furthermore, these two polymers exhibits

relatively high glass transition temperatures of >200 ^ꢀC,

which is a favorable property for devices with stable life-

times. In comparison with the well-known poly(TPD), the

electroluminescent behavior of the host materials 8e and 8f

in the prepared polymer blend PLEDs is twice as good.

7 ꢂ 10^ꢁ6cm²V^ꢁ1

s

^ꢁ1. McKeown et al.²⁸reported the hole

mobility of 8f to be 2 ꢂ 10^ꢁ6cm²V^ꢁ1s^ꢁ1and of 8a to be 8

ꢂ 10^ꢁ6cm²V^ꢁ1

s

^ꢁ1. In addition, the mobility of tert-BuPBD

in a polystyrene host is in the range of 4 ꢂ 10^ꢁ7to 9 ꢂ

10^ꢁ7cm²V^ꢁ1s^{ꢁ1 40}

Based on our PLED device characteris-

.

tics, we conclude that a hole mobility in the range of the pol-

y(TPD) derivative is a limiting factor for the electronic device

properties of those established blend systems. The polymers

with hole mobilities in the range of 10^ꢁ6cm²V^ꢁ1s^ꢁ1(side

functionalities carbazolyl, phenothiazinyl) seemed to be the

better choice for efficient and bright PLED blend systems

with tert-BuPBD and Ir(Me-ppy)₃, in that case. For an effi-

cient single-layer PLED, it is of crucial importance that the

charge carrier mobilities for both types of charges are in a

good balance, to prevent a build-up of one charge type in the

emissive layer. In small molecule devices, this challenge can

be managed easily by a tuning of the different layer thick-

nesses. In polymer blend systems, fabricated as single layer

devices, this needs to be obtained by careful variation of the

blend composition. Figure 6(b) demonstrates for the device

with the poly(TPD) a small efficiency roll-off initiating at

higher current densities. Such an efficiency roll-off has been

attributed to a triplet–triplet annihilation process between the

phosphorescent dopant and the host materials.⁴¹For the de-

vices with the carbazole- and phenothiazine-substituted poly-

mers, 8e and 8f, respectively, the efficiency roll-off is not

observable up to 20 mA cm^ꢁ2. The better efficiency of these

polymers compared with the others is an indication for a

well-balanced carrier density in the device.

The authors thank the Federal Ministry of Education and

Research of Germany (BMBF) for financial support within

the CARO project (01 BD 0685) and (01 BD 0687). The

authors thank Steffi Kreissl for device preparation and Tat-

jana Egorov-Brening for OFET preparation.

REFERENCES AND NOTES

1 Burroughes, J. H.; Bradley, D. D. C.; Brown, A. R.; Marks,

R. N.; Mackay, K.; Friend, R. H.; Burns, P. L.; Holmes, A. B.

Nature 1990, 347, 539.

2 Hebner, T. R.; Wu, C. C.; Marcy, D.; Lu, M. H.; Sturm, J. C.

Appl Phys Lett 1998, 72, 1998.

3 de Gans, B.-J.; Duineveld, P. C.; Schubert, U. S. Adv Mater

2004, 16, 203.

4 Joo, C. W.; Jeon, S. O.; Yook, K. S.; Lee, J. Y. Org Electron

2009, 10, 372.

5 Tuomikoski, M.; Suhonen, R.; Valimaki, M.; Maaninen, T.;

¨

Maaninen, A.; Sauer, M.; Rofin, P.; Menning, M.; Heusing, S.;

Puetz, J.; Aegerter, M. A. Proc SPIE 2006, 6192, 619204.

6 Adachi, C.; Baldo, M. A.; Thompson, M. E.; Forrest, S. R. J

It has to be mentioned that the given data concerning the

electroluminescence of the prepared PLEDs do not show the

maximal accessible efficiencies and luminances. Knowing

that these parameters are strongly dependent on the emis-

sion layer thicknesses, only the data of comparable layer

thicknesses were selected for a comparison.

Appl Phys 2001, 90, 5048.

7 Kohler, A.; Bassler, H. Mater Sci Eng R 2009, 66, 71.

¨

8 Chen, X.; Liao, J.-L.; Liang, Y.; Ahmed, M. O.; Tseng, H.-E.;

Chen, S.-A. J Am Chem Soc 2003, 125, 636.

9 Sandee, A. J.; Williams, C. K.; Evans, N. R.; Davies, J. E.;

Boothby, C. E.; Kohler, A.; Friend, R. H.; Holmes, A. B. J Am

¨

CONCLUSIONS

Chem Soc 2004, 126, 7041.

A series of novel styrene-based monomers and their poly-

mers have been synthesized as new hole-transporting mate-

10 Gong, X.; Ostrowski, J. C.; Bazan, G. C.; Moses, D.; Heeger,

A. J.; Liu, M. S.; Jen, A. K.-Y. Adv Mater 2003, 15, 45.

HOLE-TRANSPORTING HOST-POLYMER SERIES, THESEN ET AL.

3429

JOURNAL OF POLYMER SCIENCE: PART A: POLYMER CHEMISTRY DOI 10.1002/POLA

11 Al Attar, H. A.; Monkman, A. P.; Tavasli, M.; Bettington, S.;

28 McKeown, N. B.; Badriya, S.; Helliwell, M.; Shkunov, M. J

Bryce, M. R. Appl Phys Lett 2005, 86, 121101.

Mater Chem 2007, 20, 2088.

12 Doi, S.; Sekine, C.; Tsubata, Y.; Ueda, M.; Noguchi, T.;

29 Debeaux, M.; Thesen, M. W.; Schneidenbach, D.; Hopf, H.;

Janietz, S.; Krueger, H.; Wedel, A.; Kowalsky, W.; Johannes,

H.-H. Adv Funct Mater 2010, 20, 399.

Ohnishi, T. J Photopolym Sci Technol 2003, 16, 303.

13 Baldo, M. A.; Thompson, M. E.; Forrest, S. R. Nature 2000,

403, 750.

30 de Mello, J. C.; Wittmann, H. F.; Friend, R. H. Adv Mater

1997, 9, 230.

14 Tong, B.; Mei, Q.; Wang, S.; Fang, Y.; Meng, Y.; Wang, B.

J Mater Chem 2008, 18, 1636.

31 Wilson, J. S.; Chawdhury, N.; Al-Mandhary, M. R. A.; You-

nus, M.; Khan, M. S.; Raithby, P. R.; Kohler, A.; Friend, R. H.

J Am Chem Soc 2001, 123, 9412.

15 De Angelis, F.; Fantacci, S.; Evans, N.; Klein, C.; Zakeerud-

din, S. M.; Moser, J.-E.; Kalyanasundaram, K.; Bolink, H. J.;

Gratzel, M.; Nazeeruddin, M. K. Inorg Chem 2007, 46, 5989.

¨

32 Pope, M.; Swenberg, C. E. Electronic Processes in Organic

Crystals and Polymers, 2nd ed.; Oxford Science Publications:

Oxford, 1999; p 32.

16 You, Y.; Park, S. Y. Dalton Trans 2009, 8, 1267.

17 Gong, X.; Lim, S.-H.; Ostrowski, J. C.; Moses, D.; Bardeen,

33 Wang, D.; Li, W.; Chu, B.; Su, Z.; Bi, D.; Zhang, D.; Zhu, J.;

C. J.; Bazan, G. C. J Appl Phys 2004, 95, 948.

Yan, F.; Chen, Y.; Tsuboi, T. Appl Phys Lett 2008, 92, 053304.

18 Tokito, S.; Suzuki, M.; Sato, F.; Kamachi, M.; Shirane, K. Org

34 Meyer, J.; Hamwi, S.; Bulow, T.; Johannes, H.-H.; Riedl, T.;

¨

Electron 2003, 4, 105.

Kowalsky, W. Appl Phys Lett 2007, 91, 113506.

19 Yang, X. H.; Neher, D. Appl Phys Lett 2004, 84, 2476.

35 Lim, J. T.; Jeong, C. H.; Lee, J. H.; Yeom, G. Y.; Jeong, H.

K.; Chai, S. Y.; Lee, I. M.; Lee, W. I. J Organomet Chem 2006,

691, 2701.

20 Yang, X.; Muller, D. C.; Neher, D.; Meerholz, K. Adv Mater

¨

2006, 18, 948.

21 Kawamura, Y.; Yanagida, S.; Forrest, S. R. J Appl Phys

36 Kondakova, M. E.; Deaton, J. C.; Pawlik, T. D.; Giesen, D. J.;

Kondakov, D. Y.; Young, R. H.; Royster, T. L., Comfort, D. L.;

Shore, J. D. J Appl Phys 2010, 107, 014515.

2002, 92, 87.

22 Park, J. J.; Park, T. J.; Jeon, W. S.; Pode, R.; Jang, J.; Kwon,

J. H.; Yu, E.-S.; Chae, M. Y. Org Electron 2009, 10, 189.

37 Shirota, Y.; Kuwabara, Y.; Inada, H.; Wakimoto, T.; Nakada,

H.; Yonemoto, Y.; Kawami, S.; Imai, K. Appl Phys Lett 1994, 65,

807.

23 Adachi, C., Nagai, K.; Tamoto, N. Appl Phys Lett 1995, 66, 2679.

24 Tokito, S.; Tanaka, H.; Noda, K.; Okada, A.; Taga, Y. Appl

38 Wu, H.; Zou, J.; Liu, F.; Wang, L.; Mikhailovsky, A.; Bazan,

Phys Lett 1997, 70, 1929.

G. C.; Yang, W.; Cao, Y. Adv Mater 2008, 20, 696.

25 Shirota, Y.; Kuwabara, Y.; Okuda, D.; Okuda, R.; Ogawa, H.;

Inada, H.; Wakimoto, T.; Nakada, H.; Yonemoto, Y.; Kawami,

S.; Imai, K. J Lumin 1997, 72-74, 985.

39 Kolosov, D.; Adamovich, V.; Djurovich, P.; Thompson, M. E.;

Adachi, C. J Am Chem Soc 2002, 124, 9945.

26 Suzuki, M.; Tokito, S.; Sato, F.; Igarashi, T.; Kondo, K.;

40 Tameev, A. R.; He, Z.; Milburn, H. W.; Kozlov, A. A.; Vanni-

Ymaguchi, T. Appl Phys Lett 2005, 86, 103507.

kov, A. V.; Puchala, A.; Rasala, D. Appl Phys Lett 2002, 81, 969.

27 Thesen, M. W.; Krueger, H.; Janietz, S.; Wedel, A.; Graf, M.

41 Baldo, M. A.; Adachi, C.; Forrest, S. R. Phys Rev B 2000, 62,

J Polym Sci Polym Chem 2010, 48, 389.

10967.

3430

INTERSCIENCE.WILEY.COM/JOURNAL/JPOLA

Article Doi

Hole-transporting host-Polymer series consisting of triphenylamine basic structures for phosphorescent polymer light-Emitting diodes

DOI: 10.1002/pola.24127

Source and publish data:

Authors:

Article abstract of DOI:10.1002/pola.24127

Full text of DOI:10.1002/pola.24127

Products guided by the article

R&D Labs maybe for 847690-46-4

Relevant to this article

Hot Product