University of Groningen Electrical characterization of polymeric charge transport ... ·...

University of Groningen

Electrical characterization of polymeric charge transport layersCraciun, Nicoleta Irina

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2011

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Craciun, N. I. (2011). Electrical characterization of polymeric charge transport layers. Groningen: s.n.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 26-06-2020

https://www.rug.nl/research/portal/en/publications/electrical-characterization-of-polymeric-charge-transport-layers(0ca17fe1-87a4-4afc-b165-84082f4607c4).html

https://www.rug.nl/research/portal/en/publications/electrical-characterization-of-polymeric-charge-transport-layers(0ca17fe1-87a4-4afc-b165-84082f4607c4).html

Electrical Characterization

Of Polymeric Charge Transport

Layers

Irina Crăciun

Electrical characterization of polymeric charge transport layers

Nicoleta Irina Crăciun

PhD thesis University of Groningen, The Netherlands

Zernike Institute PhD thesis series 2011-13 ISSN: 1570-1530 ISBN: 978-90-367-4872-8 ISBN: 978-90-367-4886-5 (digital version)

The research described in this thesis was performed in the research group

Molecular Electronics: Physics of Organic Semiconductors of the Zernike Institute for Advanced Materials at the University of Groningen, the Netherlands. This

research was financially supported by the Dutch Technology Foundation STW, project #06909.

© Irina Crăciun, Groningen, 2011. All rights reserved.

Cover design by Gert-Jan Wetzelaer

RIJKSUNIVERSITEIT GRONINGEN

ELECTRICAL CHARACTERIZATION OF

POLYMERIC CHARGE TRANSPORT LAYERS

Proefschrift

ter verkrijging van het doctoraat in de Wiskunde en Natuurwetenschappen aan de Rijksuniversiteit Groningen

op gezag van de Rector Magnificus, dr. E. Sterken, in het openbaar te verdedigen op

vrijdag 6 mei 2011 om 14:45 uur

door

Nicoleta Irina Crăciun

geboren op 12 september 1981

te Slatina, Roemenië

Promotor: Prof. dr. ir. P. W. M. Blom

Beoordelingscommissie: Prof. dr. D.M. de Leeuw Prof. dr. R. Coehoorn Prof. dr. D. Vanderzande

Table of Contents

Chapter 1 Introduction 1

1.1 Introduction

1.1 Charge transport conjugated polymers

1.2.1 The role of disorder

1.2.2 Hole transport in conjugated polymers

1.2.3 Origin of the enhanced SCLC in conjugated polymers

1.1 Summary

REFERENCES

2

7

7

10

13

19

21

Chapter 2

Device Operation Of Polymer Light-Emitting Diodes 24

2.1 Introduction

2.2 Electron transport in PPV

2.3 Device model for polymer light-emitting diodes

2.4 Polymer multilayer light-emitting diodes

2.5 Outline of the thesis

REFERENCES

25

26

29

31

34

38

Chapter 3

Diffusion-Enhanced Hole Transport In Thin Polymer

Light-Emitting Diodes 40

3.1 Introduction

3.2 Space-charge limited currents in thin PPV-based hole-only devices

3.3 Modelling of the diffusion of holes from an Ohmic contact at zero-bias into a thin film

3.4 Conclusions REFERENCES

41

42

45

51

52

Chapter 4

Universal Arrhenius temperature activated charge transport in

diodes from disordered organic semiconductors 53

4.1 Introduction

4.2 Temperature dependence of the hole mobility of organic SCL

diodes

4.3 Conclusions

REFERENCES

54

55

64

65

Chapter 5

Substituted polyfluorene based hole transport layer with

tunable solubility 66

5.1 Introduction

5.2 Bilayer PLEDs with a polyfluorene based hole transport layer

5.3 Conclusions

REFERENCES

APPENDIX

67

68

81

82

83

Chapter 6

Hysteresis-free electron currents in poly(p-phenylene vinylene)

derivatives 86

6.1 Introduction

6.2 Hysteresis in the electron currents of PPV-based conjugated

polymers

6.3 Conclusions

REFERENCES

87

88

101

102

PUBLICATIONS

SUMMARY

SAMENVATTING

ACKNOWLEDGMENTS

103 104

108 112

1

Chapter 1

Introduction

Abstract This chapter reviews the concepts and charge transport properties of

conjugated polymers that will be employed in this thesis. After an introduction on

conjugated polymers the charge transport properties in these materials are

discussed. In crystalline inorganic semiconductors the atoms are perfectly aligned

in a lattice, allowing for good orbital overlap between neighboring atoms. In such

materials charges are transported by band-like motion and the mobility is very

high. For materials with a large number of imperfections as conjugated polymers,

however, the electronic states become localized. The limiting factor for charge

transport is then hopping between these localized states sites with randomly

varying energy levels and inter-site distances.

Chapter 1

2

1.1 Introduction

The research on organic semiconductors started in the 1950s when small

organic molecules in the crystalline state started to be investigated. Molecular crystals like naphthalene and anthracene were shown to exhibit semiconducting

properties [1]. Already in the early 1960s, next to photoconductivity [2, 3], organic electroluminescent devices based on anthracene crystals were demonstrated by Pope et al. [4]. An important breakthrough came in 1977 when the first highly

conducting organic polymer, chemically doped polyacetylene, was reported [5]. The initial excitement on this new class of materials decreased in the subsequent

years because these doped conducting polymers were unstable in air, brittle, and difficult to process. However, in the middle of the 1980s the interest in undoped organic semiconductors, both polymers and small molecules, revived due to a

number of important discoveries. First, at Eastman Kodak electroluminescent devices from thin film stacks of vacuum-sublimed small molecules were realized

by Tang et al. [6]. Subsequently, field-effect transistors made from polythiophene [7] and from small conjugated oligomers [8] were reported. Then, in 1990 also electroluminescence from conjugated polymer-based diodes was discovered at

Cambridge University [9]. In the middle of the 1990s also the first photovoltaic devices based on bulk-heterojunctions in blends of polymers or polymers and

fullerenes were developed [10-13]. An advantage of conjugated polymers (CPs) is that by modification of

their chemical structure they can be made soluble in common organic solvents. As

a result polymer-based devices can be processed from solution [14] by spin coating, inkjet printing techniques or coating techniques like slot-die coating. This

enables a low-cost high-speed production via roll-to-roll processing of electronic devices. As a result of this ease of processing conjugated polymers are promising

to provide large-area, flexible, lightweight lighting systems, integrated circuits and solar cells. Conjugated polymers are organic macromolecules which have a

framework of alternating single and double carbon-carbon bonds. Single bonds are referred to as σ-bonds and are associated with a highly localized electron density in

the plane of the molecule. Double bonds contain both a σ-bond and a π-bond, where the π-bond is the overlap between pz orbitals of neighboring atoms along the conjugation path. The overlapping pz orbitals establishe a delocalization of the

electrons situated above and below the plane of the molecule. π-bands are either empty (called the Lowest Unoccupied Molecular Orbital - LUMO) or filled with

electrons (called the Highest Occupied Molecular Orbital - HOMO). The band gap

Introduction

3

of these materials determined from optical measurements is within the semiconductor range of 1–4 eV, which covers the whole range from infrared to

ultraviolet region. A typical example of a conjugated polymer is polyacetylene, which consists only of a single chain of alternating single- and double-bonds

(Figure 1-1).

The absence of an ideal 3D periodic lattice in polymer semiconductors

complicates the description of charge transport processes in terms of standard

semiconductor models. An ideal crystal has a three-dimensional architecture characterized by the infinite repetition of identical structure units in space. Its

structure can be described in terms of a lattice characterized by long-range order and strongly coupled atoms [15]. For silicon or germanium this strong coupling results in the formation of long-range delocalized energy bands separated by a

forbidden energy gap [15]. Charge carriers added to the semiconductor move in these energy bands with a relatively large mean free path. Carrier scattering

significantly affects the carrier mobility, which depends on the effective mass of electrons and the temperature. Carrier mobility in the order of 10−1 m2/Vs is

reported for pure inorganic semiconducting crystals such as silicon or germanium. In organic crystals, such as pentacene, the atoms are held together by weak van der Waals or London forces. This weak coupling between molecules results in a

narrow width for the valence and conduction bands and the band structure can be easily disrupted by introducing disorder in the system. Although organic molecular

C C C C C C C C

p-bond

s-bond

CC

CC

CC

CC

Figure 1-1. Schematic representation of the electronic bonds between carbon

atoms (above) in polyacetylene (below).

Chapter 1

4

crystals still exhibit band conduction, excitations and interactions localized on individual molecules play a predominant role. Their mobility, in the order of 10−3

m2/Vs, is significantly lower than those of their inorganic counterparts. The Su-Schrieffer-Heeger (SSH) theory describes the electronic structure of conjugated

polymers (CP) [16]. It takes account of electron-phonon coupling. Furthermore, a postulation is made in the model that the pz orbitals may not be equally spreaded along the chain, but can be slightly paired. This pz-pairing results in the common

representation of the pz electrons in alternating single and double bonds (conjugation, see figure 1-1). It also gives rise to the semiconducting behaviour of

(undoped) CP’s -they would have been metallic if the pz orbitals would have been equally spreaded. These model calculations give rise to a whole spectrum of ’quasi-particles’, both charged (net on-chain charge, e.g. by doping or charge

injection) and uncharged. The most important uncharged particle is the soliton, which is a local distortion of the conjugation. It is essentially a local reversion of

the conjugation, and is localized to several chain atoms. It can only exist in polyacetylene, because in this polymer each carbon atom in the chain is coupled to

an identical carbon atom. Therefore, it makes no difference which two neighbors have their pz orbitals paired. This is a so-called degenerate state. At the other hand, in for example poly(p-phenylen vinylene) (PPV), it matters which carbon atoms in

the chain pair together, and a soliton cannot exist. The most important charged particle is the polaron. The polaron essentially is also a local distortion of the

conjugation, but now initiated by an extra charge carrier (Figure 1-2) [17], considered to be delocalized (free) to a certain extent, as is illustrated in Figure 1-2. The polaron is the medium that in the models for CP’s causes the charge transport.

A major problem when applying the SSH model to experimental systems is

that most conjugated polymers do not have a well-ordered structural configuration

Figure 1-2. Formation of a polaron in poly-(p-phenylene), shown for the case

of doping with an electron acceptor. Picture taken from [19]

A

Introduction

5

as crystals. The conjugation of the polymer backbone is easily disrupted by

chemical or structural defects, such as chain kinks or twists. Experimentally, it has

been found that the charge carrier mobility in these materials is in the range of

10−12 −10−10 m2/Vs for polymer light-emitting diodes and 10−8 − 10−5 m2/Vs for

polymer field-effect transistors. This is orders of magnitude lower than the

mobility determined for organic crystals. Over the past decades intense research

has been carried out in order to explain the transport of charge carriers in

disordered polymer semiconductors which would justify such low mobility.

1.2 Charge transport in conjugated polymers

1.2.1 The role of disorder

A semiconducting polymer is not a perfect conjugated system, because its twisted and kinked chains and chemical defects cause conjugation breaks. Because

of their spatial and energetically disordered configuration, these systems have no translation symmetry. The concept of band conduction by free charges does not

apply. Instead, the formation of localized states is enhanced and a different theoretical approach is required. In order to participate to the transport, the charge carriers must hop between these localized states (inter- or intra-chain transitions).

This usually leads to very low carrier mobility. To overcome the energy difference between two localized states, the carriers absorb or emit phonons. This process of

phonon-induced hopping was suggested by Conwell [20] and Mott [21] in connection with metallic conduction in inorganic semiconductors. Miller and Abrahams proposed a hopping model based on a single-phonon jump rate

description [22]. The hopping rate of carriers from occupied i to unoccupied j localized donor states depends on the height of the energetic barrier Ej − Ei and the

distance Rij between the states i and j:

( )[ ]

<

>−−−=→

ji

jiBjiij

jiEforE

EforETkEE

a

Ra

1

,/exp)2exp(0 γνν (1-1)

where the prefactor ν0 is the attempt-to-hop frequency, γ is the inverse localization

length, a is the average lattice distance, and kB is the Boltzmann constant. The first exponential term from Eq. 1-1 represents the tunneling probability and the second exponential term accounts for the temperature dependence of the phonon density.

Chapter 1

6

Bässler [23] proposed in 1993 a charge transport model for disordered organic systems. He assumes that electron-phonon coupling is sufficiently weak so that the

polaronic effects can be neglected, and the hopping rates can be described by the Miller-Abrahams formalism (Eq. 1-1). The charges hop in a regular array of

hopping sites. In this way both positional disorder (fluctuation in inter-site distance) and energetic disorder (fluctuation in site-energy) are introduced. In this model, the energy distribution of localized states can be approximated by a

Gaussian function [23]:

( )

−=

2

2

22

exp2 DOSDOS

tGauss

NDOS

σε

πσ (1-2)

where Nt is the total density of sites, σDOS is the width of the Gaussian density of

states (DOS) and the energy ε is measured relative to the center of the DOS. The choice for this particular DOS shape is supported by the fact that coupling between

a charge carrier and a random distribution of static or induced dipoles leads to a Gaussian DOS function [24]. The charge transport in the Gaussian disorder model (GDM) cannot be solved analytically and therefore an alternative approach of

Monte Carlo simulations has been applied [23]. Using the hopping rate from the Miller-Abrahams formalism, the Monte Carlo simulations revealed that carriers (in

this case the electrons) with an arbitrary energy within a Gaussian DOS relax to an equilibrium level −σ2

DOS/kBT below the center of the DOS distribution. The energy level that is relevant for the transport is located at −(5/9)σ2DOS/kBT. This gives rise

to the dependence µ (T) = µ0 exp [-(2σDOS/3kBT)2]. But the hopping mobility must

depend also on the electric field since the average barrier height for energetic

uphill jumps in field direction is reduced [23]. On the basis of the Monte Carlo simulations, the charge carrier mobility is temperature- and field-dependent, and in the limit of high electric fields is given by [23]:

<Σ

−

≥Σ

Σ−

×

−= ∞

5.15.2exp

5.1exp

3

2exp

2

2

2

2

forETk

C

forETk

C

Tk

B

B

B

GDM

σ

σ

σµµ

(1-3)

Introduction

7

where µ0 is the mobility in the limit T→∞, with values between 10−6 and 10−5

m2/Vs, C is a constant that depends on the site spacing, and Σ is the degree of

positional disorder. A consequence of hopping in a Gaussian DOS is the non-Arrhenius behavior of the mobility [23].

Because the results of simulations performed within the frame of the standard Gaussian disorder model can explain the experimental results only at high fields (> 108 V/m), further improvement was necessary. A spatially correlated site-

energy distribution was considered [25], which means that the energies are correlated over a greater length than the distance between hopping sites. Due to the

correlation of energies of adjacent sites, the field dependence of the mobility extends to lower electric fields. Spatial correlations in site-energy may arise from long-range charge-dipole interactions in the material, where the disorder is

determined by the random orientations of dipole moments of nearby molecules. In this correlated disordered model (CDM), the empirical expression for the mobility

is given by [26, 27]:

Γ−

−= ∞ σ

σσµµ

eaE

TkC

Tk BB

CDM

2/3

0

2

exp5

3exp (1-4)

where C0=0.78, a is the intersite separation, and Γ= 2 for organic materials. The

main difference between GDM and CDM is the predicted temperature-dependent field dependence. The GDM and CDM based hopping models of charge carriers in a Gaussian DOS have initially been used to explain the experimentally observed

temperature- and field dependence of the hole mobility in light-emitting diodes based on poly(p-phenylene vinylene) (PPV) based derivatives.

1.2.2 Hole transport in conjugated polymers Since the discovery of electroluminescence in poly(p-phenylene vinylene)

(PPV) in 1990 [9] extensive research has been carried out to understand and develop polymer light-emitting diodes (LEDs). A typical single-layer polymer LED is presented in Figure 1-3. A thin polymer film is spin-coated from solution

on a semitransparent bottom electrode, normally indium-tin-oxide (ITO), which forms the anode. A low work-function metal (calcium or barium) is evaporated on

top of the polymer and serves as cathode.

Chapter 1

8

The main processes that govern the operation of the polymer LED are: charge injection, charge transport and recombination (Figure 1-3b) [28]. Under

forward bias, holes and electrons are injected from the anode and cathode, respectively. The charge carriers move through the polymer film and recombine.

The energy released upon recombination is emitted as a photon through the semitransparent electrode. The emitted light can be tuned from red to blue, depending on the band gap of the polymer. In order to be injected from the

electrodes, the charges must surmount or tunnel through a barrier at the polymer/electrode interface, which is determined by the position of the highest

occupied molecular orbital (HOMO or π orbital) and the lowest unoccupied molecular orbital (LUMO or π* orbital) and the position of the electrode metal work-functions. A schematic band diagram of a PPV-based LED under forward

bias using ITO as a hole injector and Ba as an electron injector is shown in Figure 1.3-b.

In order to investigate the hole mobility in PLEDs, diodes with one hole

injecting contact and one electron blocking contact, in this case Au, have been fabricated. Such a structure is called hole-only diode. As the active layer the solution processable PPV-derivative poly(2-methoxy-5-(3′,7′-dimethyloctyloxy)-p-

phenylene vinylene) (OC1C10-PPV) is used. The experimentally measured current-voltage characteristics at low electric fields of these hole-only diodes show a

quadratic behaviour [28], as shown in Figure 1-4, indicating that the current is

GLASS SUBSTRATE

ANODE

POLYMER

CATHODE

PPV

HOMO

LUMO

Ba

ITO

Recombination

Injection

Transport

n

Figure 1-3. Schematic representation of a polymer LED (a) and energy

diagram of a PPV-based LED (b).

Introduction

9

space-charge limited. In this case the current density is characterized by Child’s law [29]:

3

2

08

9

L

VJ sµεε= (1-5)

where ε0 is the permittivity of vacuum, εs is the relative dielectric constant of the semiconductor, µ is the carrier mobility and L the thickness of the device. At low voltages the J-V characteristics are well described by the SCLC given by Eq. 1-5

(line), from which the zero-field mobility µp (0) = 5×10−11 m2/Vs is determined. At

high biases the current starts to increase more rapidly with voltage and Eq. 1-5 is

no longer valid. This suggests that the mobility increases with the applied voltage. By changing the temperature, it was also found that the mobility also depends on the temperature [30].

Initially, the hole transport in PPV has been described by the combination of the SCL conduction model with a temperature and field-dependent mobility

[30]:

( ) ( )( ) ( )xFTxFxepJ p ,µ= (1-6)

0.1 1 10 10010

-4

10-3

10-2

10-1

100

101

102

n

()

OR1

R2O

0.7 µm

J (A/m

2)

Vbias (V)

Figure 1-4. J–V of OC1C10-PPV hole-only diode at room temperature.

The polymer layer thickness is 700 nm. The solid line represents the

prediction of the conventional SCLC model using Eq. 1-5.

Chapter 1

10

( ) ( )xpdx

xdF

e

s =εε 0 (1-7)

( )

+

∆−== F

TkFTF

B

p γµµ exp)0(, (1-8)

where p (x) is the density of holes at position x in the semiconductor film, µ(F = 0)

is the zero-field mobility, ∆ is the zero-field activation energy, and γ is the field dependence parameter. This empirical equation which gives the temperature and field dependence of the mobility (Eq. 1-8) was noticed for the first time in poly(N-

vinyl carbazole) by Gill in 1972 [31].

1.2.3 Origin of the enhanced SCLC in PPV-based diodes

In a space charge-limited device an increase of the applied bias gives rise to a simultaneous increase of the electric field and charge carrier density. Consequently, at high voltages, it is difficult to distinguish between the

contributions of the charge carrier density and the electric field to the mobility from the current-voltage characteristics. For the understanding of the charge

transport in polymer devices it is crucial to know whether the current is governed by the field and/or the carrier density dependence of the mobility. In the first analysis of charge transport properties of polymer SCL devices [31] the charge

carrier density dependence of the mobility had not been taken into account. Neglecting the density dependence of the mobility can lead to an incorrect charge

carrier and field distribution in organic SCL diodes. In order to investigate the contribution of the charge carrier density to the mobility field-effect transistors

have been used. The basic idea of a field-effect transistor is to modulate the current that flows between two ohmic contacts, the source and the drain electrodes, by applying a voltage to a third contact, the gate electrode. The gate electrode is

electrically decoupled from the semiconductor by an insulating layer. By applying positive or negative gate voltages, induced charge carriers electrostatically

accumulate or deplete in the semiconductor close to the semiconductor/insulator interface giving rise to band bending in the semiconductor. In this way the field-effect current is varied in the source-drain channel. In contrast with conventional

monocrystalline silicon, the transport properties of disordered organic

Introduction

11

semiconductors are dominated by localized states. By applying a gate voltage in a field-effect transistor, the charge carriers accumulate in the organic semiconductor

close to the insulator, thereby filling the lower localized states. Any additional charges in the system will occupy states at relatively high energies, which means

that they will need less activation energy to jump to other sites. As a result the mobility will be enhanced and is expected to increase with charge carrier density [32,33].

In order to explain the charge transport in polymer FETs a model based on variable range hopping was developed by Vissenberg and Matters [32]. The

variable range hopping model suggests that charge carriers may either hop over a small distance with a high activation energy or hop over a long distance with a low activation energy. This transport model takes into account the filling of localized

states with charge carriers in contrast to the one developed by Bässler, which is a one particle model [23]. The model predicts that at low carrier densities and low

temperatures, the transport properties are determined by the tail states of Gaussian DOS, which is approximated by an exponential DOS [32]:

=

0

exp expTkTk

NDOS

BB

t

on

ε (1-9)

where Nt is the number of states per volume unit, T0 is the width of the exponential DOS, and є the level energy. It is considered that the energy distribution of the

carriers at equilibrium is given by the Fermi-Dirac distribution. Using percolation theory [34], an expression for the conductivity can be derived as a function of the

occupation fraction δ and the temperature T [32]:

( )( )

( )

TT

c

t

B

T

TTTN

T

/

3

0

4

0

0

0

2

sin/

,

=α

πδ

σδσ (1-10)

where σ0 is a prefactor for the conductivity, and Bc is the critical number for the onset of percolation. For three-dimensional amorphous systems Bc=1.8 [35]. The

conductivity expressed by Eq. 1-10 has an Arrhenius-like temperature dependence σ~ exp [−Ea/kBT] explained by the fact that in an exponential DOS the hopping can be described in terms of activation from the Fermi level to a specific transport level

[36]. Moreover, the conductivity increases superlinearly with the charge carrier

Chapter 1

12

density σ~δT0/T. This is due to the fact that by increasing the carrier density the

states are filled and an activated jump to the transport energy is facilitated.

From electrical measurements using as active material OC1C10-PPV in a FET geometry a field-effect mobility of 4.7 × 10−8 m2/Vs at Vg = −19 V at room

temperature has been obtained [37] . Surprisingly, this value for the field-effect mobility is approximately three orders of magnitude larger than the mobility value determined from a hole-only diode, as shown in Figure 1-4. The question is why

the experimental mobilities for the same polymer differ when measured in different device geometries (LED and FET). The answer is related to the charge carrier

density regime in which the two devices operate. For a 700 nm film of OC1C10-PPV for the voltage range applied of 1 V to 10 V (Figure 1-4) the mobility is charge carrier and field-independent. This voltage range corresponds to hole

density of 2.5 × 1020 to 2.5 × 1021 m−3. [37] In FETs, the charge carrier density is much larger at the semiconductor/insulator interface, ranging from ~1023 m−3 at low

gate voltage (-1 V) to 3 × 1025 m−3 at a gate bias of -20 V. The experimental mobilities obtained from the hole-only diode and the field-effect transistors are

presented together in Figure 1-5 for OC1C10-PPV. The combined results from the diode and field-effect measurements show that the hole mobility is constant for charge carrier densities < 1022 m−3 and increases with a power law for densities >

1022 m−3. The big differences in mobility values obtained from diodes and FETs, based on a single semiconducting disordered polymer, are due to the large

differences in charge densities.

Figure 1-5. Hole mobility as function of hole density p in hole-only

diode and FET for OC1C10-PPV (symbols) [37]. The dashed lines is a

guide for the eye.

1021

1022

1023

1024

1025

10-10

10-9

10-8

10-7

T=293 K

µ LE

D,

µ FE

T (

m2/V

s)

p (m-3)

Introduction

13

0.1 1 1010

-6

10-4

10-2

100

102

T=293 K

T=275 K

T=255 K

T=235 K

J (A/m

2)

V (V)

Figure 1-6. Temperature dependent J–V characteristics of the OC1C10-PPV

hole-only diode. The solid lines represent the prediction from Eq. 1-11.

Combination of the diode and field-effect measurements shows that at room temperature the dependence of the hole mobility on charge carrier density can be

described by the empirical relation [37]:

(1-11)

where µh(0,T) is the hole mobility at low densities obtained from the quadratic SCL current. It should be noted that this unification is only possible for highly

disordered polymers such as OC1C10-PPV, in which the charge transport is isotropic [37].

Now knowing how the mobility depends on density the SCL current was

calculated using only the density-dependent mobility as given by Eq. 1-11 [38]. By numerically solving Eqs. 1-6, 1-7 and 1-11, the J-V characteristics of OC1C10-PPV

hole-only diodes are obtained, as shown in Figure 1-6 by the solid lines. It can be observed that at T = 275 K and T = 293 K the calculated SCL

current density for OC1C10-PPV at high fields is in good agreement with the

( ),

2

sin

),0(),(1/

/

3

0

4

0

0 0

0

−

+= TT

TT

c

hh pB

T

T

T

T

eTTp

α

πσ

µµ

Chapter 1

14

experimental current densities. Furthermore, from Figure 1-6 it appears that at temperatures lower than T = 275 K the carrier density dependence of the mobility

alone does not explain the observed increase of the SCLC. At low temperatures the field dependence becomes more important. This is due to the fact that at low

temperatures the activated hops between neighboring sites are strongly suppressed, which means that the charge transport is also suppressed. Application of an electric field leads to a reduction of these dominant barriers for the charge transport in the

field direction, resulting in strong field dependence. Pasveer et al [39] have presented the Extended Gaussian Disorder Model (EGDM) of the mobility that is

both field- and charge carrier concentration dependent. According to this model, that has also been experimentally verified, the carrier concentration dependence is dominating at room temperature when the field is not very high. At low

temperature or at high field the electric field dependence becomes important. The Pasveer mobility is given by:

),(),(),,( ETpTEpT Ep µµµ = (1-12a)

where

( )

−

+

−=

δσσσµµ 3

22

0 22

142.0exp),( pa

TkTkTkpT

BBB

p (1-12b)

and

−

+

−

= 18.012.244.0exp),(

22/3

σσ

µEea

TkET

B

E (1-12c)

The exponent δ in (3.10b) is given by

( )( ) ( )( )22

/

4lnln/ln2

Tk

TkTk

B

BB

σ

σσδ

−−= (1-12d)

At low carrier concentrations the average energy of the charge carriers is

given by the equilibrium energy -σ2/kT. The charge carriers only occupy the tail

states of the Gaussian DOS and the average distance between charge carriers is

sufficiently large that the transport of a carrier is not affected by the presence of other carriers. As a result the mobility is constant at low densities. Above a certain

critical concentration the Fermi-level will pass the equilibrium level and the

Introduction

15

average energy of the charges will increase substantially with increasing concentration [40]. As a result the energy required to hop to neighboring sites will

decrease, leading to a mobility that increases with charge carrier density. Note that the carrier concentration becomes more important as the disorder in the material is

increased. The domination of the density dependence at room temperature has also been verified by studying the thickness dependence of PPV-based SCL hole-only devices [41]. The EGDM model has recently been extended by also taking into

account spatial correlations between the site energies. In recent studies, however, in two commonly used classes of conjugated polymers no indications for

correlation effects were found [42, 43]. It should be noted that in these charge transport models the formation of polarons as sketched in figure 1-3 is neglected. When polaronic effects are dominating the localization of charge carriers instead of

disorder hopping rates following Marcus theory [44] would be more appropriate. However, in that case an increase of the mobility with charge carrier density is

highly unlikely [45].

1.3 Summary

Since the discovery of electroluminescence in conjugated polymers it has been recognized that charge transport is a key ingredient for the efficiency of the

polymer light-emitting diodes (PLEDs). From temperature dependent current-density vs. voltage characteristics it has been obtained that the hole transport in poly(dialkoxy-p-phenylene vinylene) (PPV) is governed by the dependence of the

hole mobility on electric field and charge carrier density. At room temperature the charge carrier density dependence of the mobility is dominant, while at low

temperatures the field dependence of the mobility must be taken into account. The origin of the hole mobility is generic for a large class of disordered materials, and arises from hopping in a system with both energetic and structural disorder. In the

next chapter an overview of the device operation of polymer based LEDs is given.

Chapter 1

16

REFERENCES:

[1] K. C. Kao and W. Hwang, Electrical transport in solids, with Particular Reference to Organic Semiconductors (Pergamon, Oxford, 1981); M. Pope

and C. E. Swenberg, Electronic Processes in Organic Crystals (Clarendon, Oxford, 1982).

[2] O. H. Le Blanc, J. Chem. Phys. 33 (1960) 626.

[3] K. G. Kepler, Phys. Rev. 199 (1960) 1226. [4] M. Pope, H. P. Kallmann, and P. Magnante, J. Chem. Phys. 38 (1963)

2042. [5] C. K. Chiang, C. R. Fincher, Jr., Y. W. Park, A. J. Heeger, H. Shirakawa,

E. J. Louis, S. C. Gau, and A. G. MacDiarmid, Phys. Rev. Lett. 39 (1977)

1098. [6] C. W. Tang and S. A. VanSlyke, Appl. Phys. Lett, 51 (1987) 913.

[7] H. Koezuka, A. Tsumura, Y. Ando, Synth. Met. 18 (1987) 699; A. Tsumara, H. Koezuka, and Y. Ando, Synth. Met. 25 (1988) 11.

[8] G. Horowitz, D. Fichou, X. Z. Peng, Z. Xu, and F. Garnier, Solid State

Commun. 72 (1989) 381; F. Garnier, G. Horowitz, X. Z. Peng, and D. Fichou, Adv. Mater. 2 (1990) 592.

[9] J. H. Burroughes, D. D. C. Bradley, A. R. Brown, R. N. Marks, K. Mackey, R. H. Friend, P. L. Burn, and A. B. Holmes, Nature 347 (1990) 539.

[10] C. W. Tang, Appl. Phys. Lett. 1986, 48, 183. [11] J. J. M. Halls, C. A. Walsh, N. C. Greenham, E. A. Marseglia, R. H.

Friend, S. C. Moratti, A. B. Holmes, Nature 1995, 376, 498. [12] G. Yu, J. Gao, J. C. Hummelen, F. Wudl, A. J. Heeger, Science 1995, 270,

1789. [13] C. J. Brabec, N. S. Sariciftci, J. C. Hummelen, Adv. Funct. Mater. 2001,

11, 15.

[14] D. Braun, and A. J. Heeger, Appl. Phys. Lett. 58 (1991) 1982. [15] C. Kittel, Introduction to solid state physics, 7th Ed., John Wiley and Son,

Inc. 1996. [16] W. P. Su, J. R. Schrieffer and A. J. Heeger, Phys. Rev. Lett. 1979, 42,

1698.

[17] J. L. Bredas and G. B. Street, Accounts of Chemical Research 1985, 18, 309.

[18] K. C. Kao and W. Hwang, Electrical transport in solids with particular reference to organic semiconductors.Pergamon Press, 1981.

Introduction

17

[19] P.F. van Hutten et al., Opto-Electronic Properties of Polymerst. lecture notes (text&images) for a course at university of Groningen.

[20] E. M. Conwell, Phys. Rev. 1956, 103, 51. [21] N. F. Mott, Can. J. Phys. 1956, 34, 1356.

[22] A. Miller, E. Abrahams, Phys. Rev. 1960, 120, 745. [23] H. Bässler, Phys. Stat. Sol. B 1993, 175, 15. [24] A. Dieckmann, H. Bässler, P. M. Borsenberger, J. Chem. Phys. 1993, 99,

8136. [25] Yu. N. Gartstein, E. M. Conwell, Chem. Phys. Lett. 1995, 245, 351.

[26] D. H. Dunlap, P. E. Parris, V. M. Kenkre, Phys. Rev. Let. 1996, 77, 542. [27] S. V. Novikov, D. H. Dunlap, V. M. Kenkre, P. E. Parris, A. V. Vannikov,

Phys. Rev. Let. 1998, 81, 4472.

[28] P. W. M. Blom, M. C. J. M. Vissenberg, Mat. Sci. Eng. 2000,27, 53. [29] M. A. Lampert, P. Mark, Current injection in solids, Academic Press,

New York 1970. [30] P. W. M. Blom, M. J. M. de Jong, M. G. Van Munster, Phys. Rev. B 1997,

55, R656. [31] W. D. Gill, J. Appl. Phys. 1997, 43, 5033. [32] M. C. J. M. Vissenberg, M. Matters, Phys. Rev. B 1998, 57, 12964.

[33] A. R. Brown, C. P. Jarrett, D. M. de Leeuw, M. Matters, Synth. Met. 1997, 88, 37.

[34] V. Ambegaokar, B. I. Halperin, J. S. Langer, Phys. Rev. B 1971, 4, 2612 [35] G. E. Pike, C. H. Seager, Phys. Rev. B 1974,10, 1421. [36] D. Monroe, Phys. Rev. Lett. 1985,54, 146.

[37] C. Tanase, E. J. Meijer, P. W. M. Blom, D. M. de Leeuw, Phys. Rev.Lett. 2003, 91, 216601.

[38] C. Tanase, P. W. M. Blom, D. M. de Leeuw, Phys. Rev. B 2004, 70, 1. [39] W. F. Pasveer, J. Cottaar, C. Tanase, R. Coehoorn, P. A. Bobbert, P.W. M.

Blom, D. M. de Leeuw, M. A. J. Michels, Phys. Rev. Lett. 2005, 94,

206601. [40] R. Coehoorn, W. F. Pasveer, P. A. Bobbert, and M. A. J. Michels, Phys.

Rev. B 72, 155206 (2005). [41] P. W. M. Blom, C. Tanase, D. M. de Leeuw, R. Coehoorn, Appl. Phys.Lett.

2005, 86, 92105.

[42] M. Bouhassoune, S. L. M. van Mensfoort, P. A. Bobbert, and R. Coehoorn, Org. Electr. 10 437 (2009).

[43] R. J. de Vries, S. L. M. van Mensfoort, V. Shabro, S. I. E. Vulto, R. A. J. Janssen, and R. Coehoorn, Appl. Phys. Lett. 94, 163307 (2009).

[44] R. A. Marcus, Rev. Mod. Phys. 65, 599 (1993).

Chapter 1

18

[45] I. I. Fishchuk, V. I. Arkhipov, A. Kadashchuk, P. Heremans, and H. Bässler, Phys. Rev. B 76, 045210 (2007).

19

Chapter 2

Device Operation Of Polymer

Light-Emitting Diodes

Abstract This chapter reviews a device model for the current and light generation of

polymer light-emitting diodes (PLEDs). The model is based on experiments carried out on poly(dialkoxy-p-phenylene vinylene) (PPV) devices. The electron

conduction in PPV is strongly reduced by the presence of traps with a total density of 1018cm-3. The unbalanced electron and hole transport gives rise to a bias-dependent efficiency due to quenching of excitons at the metallic cathode. This

efficiency reduction can be circumvented using multilayers, where the electroluminescence is confined to a single emission layer. A major problem for

polymer based multilayer devices is the solubility of the materials used; a multilayer can not be fabricated when a spin casted layer dissolves in the solvent of the subsequent layer. An overview is given of various approaches to realize

multilayer devices, followed by an outlook of the thesis.

Chapter 2

20

2.1 Introduction

Organic light-emitting diodes (OLEDs) are considered as promising candidates for lighting applications and full-color emissive displays. Organic LEDs

are typically made from either small molecules [1] or from conjugated polymers [2]. Small molecule based devices are deposited with vacuum techniques, whereas soluble conjugated polymers are processed from solution. Wet-processing

techniques are more suited than vacuum deposition in terms of roll-to-roll processing of organic electroluminescent materials onto flexible plastic substrates.

An advantage of evaporated small molecule based devices is that the active part consists of various layers with various functions, leading to highly efficient devices [3]. These layers are chosen to have properties such as hole and electron transport,

hole or electron blockage and high emission. However, the preparation of multilayers from solution is problematic because the bottom layer can be dissolved

during application of a subsequent layer. A typical PLED consists of a thin layer of undoped conjugated polymer

sandwiched between two electrodes on top of a glass substrate. In the early days attention had especially been focused on PLEDs that contain the conjugated polymer poly(phenylene vinylene) (PPV) [2] or its derivatives. The PPV is spin-

coated on top of a patterned indium-tin-oxide (ITO) bottom electrode, which forms the anode. The cathode on top of the polymer consists of an evaporated metal layer

for which Ca is used, as was shown in Figure 1-3.

2.2 Electron transport in PPV In the previous chapter an overview has been given on the transport of holes in poly(2-methoxy-5-(3′,7′-dimethyloctyloxy)-p-phenylene vinylene) (OC1C10-PPV).

However, it is evident that understanding of the transport of electrons is also essential for the optimization of double-carrier devices such as polymer light-emitting diodes. Without an appropriate description of the electron transport the

electrical characteristics of these double carrier devices cannot be described. Experimentally, it has been shown that the electron current in poly (dialkoxy-p-

phenylene vinylene) (PPV) derivatives is strongly reduced as compared to the hole current [4,5]. The strongly reduced electron current was explained by trap-limited conduction [4]. The strong dependence of the electron current on applied voltage

Device Operation Of Polymer Light-Emitting Diodes

21

and sample thickness was in accordance with an exponential distribution of trapping sites in energy, given by (E<Ec )

−

=

tB

c

tB

t

tTk

EE

Tk

NEn exp)( , (2-1)

with nt (E) the trap density of states at energy E, Ec the energy of the conduction

band, Nt the total density of traps, and kTt an energy characterizing the trap distribution. The J-V characteristics following from the exponential trap

distribution (2-1) are given by [6]

)(12

10 rC

L

V

eNeNJ

r

rr

t

r

nc +

+

=

εεµ (2-2)

with r=Tt /T, the effective density of states in the conduction band estimated as Nc

=2.5×1019 cm-3, and C(r)=rr(2r+1)

r+1(r+1)

-r-2 .

Figure 2-1. Thickness dependent J–V characteristics for devices between 200 nm

and 440 nm (symbols) and the simulation for the respective thicknesses (lines),

using the exponential trap model, with Nt=8.5×1023 m

–3 and Tt=1500 K, and a

zero-field mobility of 5×10-11 m

2/Vs [7].

Chapter 2

22

The J-V characteristics according to Eq. (2-2) are then determined by three parameters: the carrier mobility µn , the trap density Nt , and the trap distribution

parameter Tt . The latter follows directly from the slope of the logJ vs logV characteristic according to Eq. (2.2). For poly(2-methoxy-5-(3′,7′-

dimethyloctyloxy)-p-phenylene vinylene) (OC1C10-PPV) this yields Tt=1500 K.

Assuming µp=µn the amount of traps is modeled to be Nt=1×1018 cm-3. Using these

parameters a good agreement between experimental and calculated results has been

obtained [7], as shown in Figure 2-1. However, it was demonstrated that trap-limited electron currents exhibited

a much weaker temperature dependence than expected [8]: the exponential trap model, as derived by Mark and Helfrich [9] within the band transport formalism and given by Eq. 2-2, could not explain this reduced temperature dependence.

As discussed in Chapter 1 in case of disordered organic semiconductors like conjugated polymers, the transport of free charges has been described by

hopping in a Gaussian DOS. Instead of a sharply defined conduction band edge, there is a broad distribution of states of the lowest unoccupied molecular orbital (LUMO) through which the transport occurs, as schematically shown in Figure 2-

2. DOS

EE0EF

E0-E

a

Etc-E

a

Figure 2-2. (a) The exponential trap distribution in case of a disordered system

with a Gaussian DOS centered around E0. The Fermi Energy EF is supposed to lie

below the center of the trap DOS [8].

Re-evaluation of the Mark-Helfrich equation (2-2) by taking into account the influence of this Gaussian distribution of the LUMO leads to a current-voltage

characteristic of the form [8]


23

12

11

/)(

0

11

12+

++

−

+

++

=

r

rrrr

kTEE

t

rec

L

V

r

r

r

r

eqNqNJ

tatc

εεµ (2-3)

Compared to Equation 2-2 the energy shift Ea , given by σ2/2KT, now mimics the

role of the conduction band edge relative to E0, which for this case is, however,

temperature dependent. This temperature dependence can be taken into account by incorporating Ea in the effective concentration of traps

( )tatctefft kTEENN /)exp)(

−= , (2-4)

varying with temperature as [ ]tkT//kT))1/2((exp 2σ− . The total concentration of

traps Nt is a constant value and can only be determined if the energy position of the

distribution Etc is known. This modified trap model consistently describes the reduced temperature dependence as observed in the experiments on electron transport in three PPV-derivatives [8]. 2.3 Device model for polymer light-emitting diodes

For PLEDs, in which both electrons and holes are injected, the different conduction

of electrons and holes is directly responsible for the distribution of the light-output in the polymer layer. The first generation PLED device model was based on the

theory of conventional double-carrier semiconductor devices [11]. For a double-carrier device two additional phenomena are of importance, namely recombination and charge neutralization. Owing to charge neutralization the total charge (holes

plus electrons) may far exceed the net charge (holes minus electrons). As a result the current density in a double-carrier device can be considerably larger than in a

single-carrier device. In the simple case without traps and a field- and density independent mobility the double carrier current (plasma limit) is given by [11]

3

22/1

0

0

2/1 )(2

8

9

L

V

B

eJ

r

npnp

r

+

=εε

µµµµεε

π , (2-5)

Chapter 2

24

with B the bimolecular recombination constant. Including electron trapping and a field- and density dependent mobility the PLED device model is chararacterized by

the current-flow equation

)()()](),([)()()](),([ xExnxnxEexExpxpxEeJJJ nppn µµ +=+= ,

(2-6) the Poisson equation

)()()()(0 xnxnxp

dx

xdE

et

r −−=εε

, (2-7)

and the particle-conservation equations

)()(11

xnxBpdx

dJ

edx

dJ

e

pn =−= . (2-8)

In the above equations, p(x) and n(x) represent the density of mobile holes

and electrons, nt(x) the density of trapped electrons, and E(x) the electric field as a

function of position x. The electron trap distribution Nt(E) as a function of energy E is assumed to be exponential with a characteristic energy kTt, according to Eq. (2-

1). The set of Eqs. (2-6)-(2-8) can be solved numerically. Model calculations of a PLED with Ohmic contacts showed that the light-output is mainly confined in a

region close to the cathode, due to the reduced electron conduction [10]. As a result non-radiative energy transfer to the metallic cathode strongly reduces the quantum efficiency QE (photon/charge carrier) of the PLED at low voltages. At higher

voltages the difference between electron- and hole transport decreases and the recombination zone shifts away from the interface, resulting in an increase of the

quantum efficiency [10]. In a PLED the recombination strength between the electrons and holes is of the Langevin type [10]. As a result both light-output and current are proportional to the charge carrier mobility. Consequently, the quantum

efficiecy QE of a PLED, which represents light-output/current, is expected to be independent of the mobility. It was observed that for PPV-derivatives with higher

mobilities an increase of the charge carrier mobility is accompanied by an enhancement of the width of the exciton quenching region close to the metallic cathode [12]. The resulting decrease of the quantum efficiency gives rise to a

reduction of the power efficiency and to a distinct optimum for the charge carrier mobility. The optimum mobility of typically 10-11 m2/Vs is the best compromise

between an enhanced charge transport, reduction of the luminescence efficiency


25

and enhanced quenching at the cathode. A different route to enhance the efficiency of PLEDs is using multilayer stacks.

2.4 Polymer multilayer light-emitting diodes

As mentioned in the previous paragraph the use of only a single electro-optic layer

has a large fundamental disadvantage: due to the reduced electron transport in conjugated polymers most of the light in a PLED is generated close to the metallic

cathode. This metallic cathode acts as a quenching site for the generated excitons, thereby strongly reducing the efficiency of the PLEDs. These fundamental limitations can be circumvented by using devices consisting of a number of active

layers. In such a heterojuction device electrons and holes can not leave the device without recombining, as has been demonstrated for OLEDs based on evaporated

small molecules with a nearly 100% internal quantum efficiency [13]. A major problem for solution processed materials is the integrity of the stack; a multilayer

can not be fabricated when a spin casted layer dissolves in the solvent of the subsequent layer.

In recent years a number of approaches have been developed to realize

solution processed multilayer PLEDs. In a first approach efficient bi-layer have been realized using a precursor polyphenylene vinylene (PPV) as a hole transport

layer (HTL) [14]. In another approach orthogonal solvents are used for the construction of organic multilayers: By subsequently depositing materials that can be dissolved in polar and non-polar solvents a three layer polymer LED was

fabricated [15]. The success of the approach of using subsequently polar- and non-polar solvents has the disadvantage that it requires a lot of synthetical effort to

adjust the solubility of the materials while maintaining good optical and electrical properties. An alternative approach is the use of cross-linkable materials in order to prevent re-dissolving during deposition of subsequent layers [16-18]. Crosslinking

is typically induced by ultraviolet (UV) exposure, comparable to standard photoresist. In order to enable crosslinking it is necessary to have

monomers/prepolymers with reactive groups in the polymer film. These groups, however, can have a negative effect on the charge transport and luminescent properties of the polymer. For crosslinking the challenge is to achieve long lifetime

of the devices, since excess material from the crosslinking process may remain in the device. Another technique for obtaining multilayer structures is to use a liquid

buffer layer during processing, as reported for a polymer-based multilayer OLED [19]. The buffer layer of (1,2-propylene glycol), that does not dissolve in the

solvents of the polymers, is deposited on top of the first polymer layer. The second

Chapter 2

26

polymer layer then initially floats on the buffer layer until the buffer layer evaporates either during the deposition process or a baking step. A challenge here

is that some (1,2-propylene glycol) may remain in the device structure and degrade performance. As a final approach the use of poly(2,7-(9,9-di-n-octylfluorene)-alt-

(1,4-phenylene-((4-sec-butylphenyl)imino)-1,4-phenylene)) (TFB) as an interlayer between PEDOT and the light-emitting layer is reported. The TFB layer is hard baked at temperatures in the range of 180oC before depositing the second polymer

layer from an organic solvent [20]. It has been seen that for a TFB layer that was originally 70 nm thick only 8-10 nm has been lost after it has been in contact with

the second solvent-based layer [21]. Another approach is to tune the solubility by chemical modification [22]:

copolymers with selective solubility can be achieved without loss of the enhanced

charged transport properties. It was shown that by shortening the (2’-ethylhexyloxy) side chains, from poly[2,5-bis(2’ethylhexyloxy)-1,4-

phenylenevinylene] (BEH-PPV), to butoxy side chains the polymer poly[2,5-bis(butoxy)-1,4-phenylenevinylene] (BB-PPV) was obtained, which is only soluble

in chloroform in very low concentrations. Consequently, by tuning the ratio of the BEH- and BB- monomers the solubility could be adjusted over the whole spectrum of solvents, thereby preserving the enhanced charge transport properties. For

example, a HTL of a BEH-BB-PPV copolymer in 1:3 ratio was only soluble in chloroform, making it compatible with a large number of light-emitting polymers.

However, a disadvantage of these PPV-based bilayer PLEDs is that the HOMO and LUMO levels of the HTL and light-emitting polymer are aligned, such that the electrons were not blocked at the interface.

2.5 Outline of the thesis

In an improved double-layer device the HTL should have a larger bandgap

than the light emitting polymer layer (LEP). At the same time, the HOMO levels of

these two polymers have to align in order to efficiently inject the holes from the HTL into the LEP, as schematically shown in Figure 2-3. Furthermore, the

chemical structure of the HTL has to be designed in such a way that when the LEP layer is spin-coated on top it will not be dissolved.


27

Figure 2-3. Schematic representation of a dual layer PLED with a hole transport

layer (HTL) that blocks electrons and a light-emitting polymer (LEP). The dual

layer PLED is sandwiched between an Ohmic PEDOT:PSS contact and a Ba/Al

cathode.

In order to limit the voltage drop across the HTL it should be kept as thin

as possible. In chapter 3 it is demonstrated that for very thin films, with

thicknesses smaller than 100 nm, the current is strongly enhanced as compared to the expected space-charge limited current. Applying the standard SCL model to

measurements on a hole-only diode with a thickness of only 40 nm results in an apparent increase of the hole mobility of a factor of 40. We demonstrate that this strong increase of the hole transport properties in these thin devices originates from

the presence of an Ohmic hole contact. For Fermi-level alignment, holes diffuse from the contact into the MEH-PPV, forming an accumulation layer with a width

of a few tens of nanometers. Due to the density dependence of the mobility, the hole transport in this accumulation region is strongly enhanced. For the analysis of

thin PLEDs, it is therefore essential that both drift and diffusion of charge carriers are taken into account. The presence of these background charges from the contact also influences the temperature dependence of the charge transport in SCL diodes.

As discussed in chapter 1 charge transport models developed for disordered organic semiconductors predict a non-Arrhenius temperature dependence

ln(µ)∝1/T2 for the mobility. In chapter 4 it is demonstrated that in space-charge

Chapter 2

28

limited diodes the hole mobility µh of a large variety of organic semiconductors shows a universal Arrhenius temperature dependence µh(T)=µ0exp(-∆/kT) at low

fields, due to the presence of extrinsic carriers from the Ohmic contact. The transport in a range of organic semiconductors, with a variation in room

temperature mobility of more than 6 orders of magnitude, is characterized by a universal mobility µ0 of 30–40 cm

2/Vs. As a result, we can predict the full temperature dependence of their charge transport properties with only the mobility

at one temperature known. In chapter 5 we report the development of wide band gap poly(9,9-

dioctylfluorene (PFO)-triarylamine based hole transport layers (HTLs) with a tunable solubility. The HOMO of this HTL aligns with the HOMO of PPV-based light-emitting polymers, which enable the construction of a heterojunction with

electron blocking functionality, as shown in Fig. 2-3. The solubility of the HTL can be tuned by adjustment of the chemical structure, without loss of the charge

transport properties. Double-layer polymer light-emitting diodes are constructed with a HTL that is not soluble in toluene at room temperature and as light-emitting

layer. The addition of the HTL enhances the efficiency of the PLED with 10% at higher voltages. In order to further enhance the efficiency and to prevent quenching of the excitons

at the cathode the PLED structure should be extended to a three layer device as sketched schematically in Figure 2-4.

Figure 2-4. Schematic band diagram of a three layer PLED. The active part of this three-layer device consists of a hole transport layer, a

luminescent layer, and an electron transport layer. In this device all the recombination is confined to the luminescent layer, thereby enabling high

efficiencies. The electron currents in many conjugated polymers, as explained in section 2-2, are low due the presence of electron traps. A major problem with the

Ca

Hole

transport

Electron

transport


29

investigation of the electron transport in the ETL is that the resulting current density vs. voltage (J-V) characteristics of these devices often exhibit strong

hysteresis effects. In chapter 6 the origin of this hysteresis is reported for electron-only devices based on derivatives of poly(p-phenylene vinylene) (PPV) varying the

hole-blocking bottom electrodes as well as the purification of the polymer. The use of a variety of hole blocking bottom contacts, as metallic electrodes and n-type doped polymers, did not give any improvement in the observed hysteresis. By

purification of the PPV, hysteresis free electron-only currents can be obtained. The deep traps responsible for hysteresis, with a concentration in the 1016 cm-3 range,

are not responsible for the trap-limited electron transport as observed in purified PPV.

Chapter 2

30

REFERENCES:

[1] C.W. Tang and S.A. VanSlyke, Appl. Phys. Lett. 51, 913 (1987). [2] J. H. Burroughes, D. D. C. Bradley, A. R. Brown, R. N. Marks, K.

Mackey, R. H. Friend, P. L. Burn, and A. B. Holmes, Nature 347, 539

(1990). [3] B.W. D’Andrade, M.A. Baldo, C. Adachi, J. Brooks, M.E. Thompson, and

S.R. Forrest, Appl. Phys. Lett. 79, 1045 (2001). [4] P. W. M. Blom, M. J. M. de Jong, J. J. M. Vleggaar, Appl. Phys. Lett. 68,

3308 (1996).

[5] L. Bozano, S. A. Carter, J. C. Scott, G. G. Malliaras and P. J. Brock, Appl. Phys. Lett. 74, 1132 (1999).

[6] K. C. Kao, W. Hwang, Electrical transport in solids (Pergamon Press, 1981).

[7] M. M. Mandoc, B. de Boer, P. W. M. Blom, Phys. Rev. B 73, 155205

(2006). [8] M. M. Mandoc, B. de Boer, G. Paasch, P. W. M. Blom, Physical Review B

75, 193202 (2007). [9] P. Mark and W. Helfrich, J. Appl. Phys. 33, 205 (1962). [10] P. W. M. Blom and M. J. M. de Jong, IEEE J. Sel. Top. Quant. El. 4, 1077

(1998). [11] M. A. Lampert, P. Mark, Current injection in solids, Academic Press, New

York 1970. [12] P.W.M. Blom, M.C.J.M. Vissenberg, J.N. Huiberts, H.C.F. Martens, and

H.F.M. Schoo, Appl. Phys. Lett. 77, 2057 (2000).

[13] S. Reineke et al., Nature 459, 234 (2009). [14] N. C. Greenham, S.C. Moratti, D. D. Bradley, R. H. Friend, A. B. Holmes,

Nature 365, 628 (1993). [15] X. Gong, S. Wang, D. Moses, G.C. Bazan, and A.J. Heeger, Adv. Mater.

17, 2053 (2005). [16] B. Domercq, R.D. Hreha, Y.D. Zhang, A. Haldi, S. Barlow, S. Marder, B.

Kippelen, J. Pol. Science: Part B: Polymer Physics 41, 2726 (2003).

[17] R-Q. Png, P-J. Chia, S. Sivaramakrishnan, L-Y. Wong, M. Zhou, L-L. Chua, and P. K-H. Ho, Appl. Phys. Lett. 91, 013511 (2007).

[18] N. Rehmann, D. Hertel, K. Meerholz, H. Becker and S. Heun, Appl. Phys. Lett. 91, 103507 (2007).

[19] S-R. Tseng, H-F. Meng, C-H. Yeh, et al., Synthetic Metals 158. 130 (2008)

[20] J-S. Kim, R.H. Friend, I Grizzi and J. Burroughes, Appl. Phys. Lett. 87, 023506 (2005).


31

[21] J-J. Park, T-J Park, W-S. Jeon et al., Organic Electronics 10, 189 (2009). [22] C. Tanase, J. Wildeman and P.W.M. Blom, Adv. Funct. Mater. 15, 2011

(2005).

Chapter 2

32

33

Chapter 3

Diffusion-Enhanced Hole

Transport In Thin Polymer Light-

Emitting Diode

Abstract The transport of holes in polymer light-emitting diodes (PLEDs) based on

poly(2-methoxy, 5-(2’ ethylhexyloxy)-p-phenylene vinylene) (MEH-PPV) is investigated as a function of layer thickness. For thicknesses smaller than 100 nm,

the current in these thin PLEDs is strongly enhanced as compared to the expected space-charge limited (SCL) current. Applying the standard SCL model to measurements on a PLED with a thickness of only 40 nm results in an apparent

increase of the hole mobility of a factor of 40. We show that this strong increase of the hole transport properties in these thin devices originates from the presence of

an Ohmic hole contact. For Fermi-level alignment, holes diffuse from the contact into the MEH-PPV, forming an accumulation layer with a width of a few tens of

nanometers. Due to the density dependence of the mobility, the hole transport in this accumulation region is strongly enhanced. For the analysis of thin PLEDs, it is therefore essential that both drift and diffusion of charge carriers are taken into

account.

Chapter 3

34

3.1 Introduction

As discussed in Chapter 1 in the space-charge limited regime the carrier density and electric field are simultaneously increased upon application of a voltage. At room temperature the enhancement of the mobility at higher voltages is totally due

to the increase of the carrier density. In the classical model for SCL conduction there are two important requisites: the first is the presence of an Ohmic contact,

such that the electric field at the injecting contact is close to zero. Furthermore, the amount of injected charges scales linearly with the applied voltage. As a result it is assumed that at zero bias there is no net charge present in the device, whereas upon

application of a voltage charges are injected into the device, thereby forming a space-charge region. However, when an Ohmic contact is applied to a

semiconductor a charge accumulation region is formed close to the contact in order to align the Fermi level, with a typical width of 10-20 nm [1], as schematically

indicated in Figure 3-1. As a result for very thin layers approaching the thickness of the accumulation layer(s) this diffusion driven charge accumulation leads to a considerable amount of net background charges in the device.

In regular semiconductors the presence of this additional background

charge will mainly show up at low voltages as an Ohmic current that depends linearly on the applied voltage. At higher voltages the number of injected charges

Figure 3-1. Schematic representation of the band diagram of a thin and

thick PPV-based hole-only device. The PPV-film is sandwiched between an

Ohmic PEDOT:PSS contact and a Au contact. Also indicated are the hole

accumulation regions close to the contacts.

Diffusion-Enhanced Hole Transport In Thin polymer Light-Emitting Diodes

35

will exceed the background charge and the standard SCL current is observed. For conjugated the strong dependence of the mobility on carrier density will amplify

the effect of such an additional background density on the charge transport. In studies performed on SCL currents in conjugated polymers so far, diffusion of

charge carriers from the contacts into the device was not taken into account. Consequently, the formation of an accumulation region with high carrier density close to the contact was ignored. An important question is now whether the

omission of this effect is relevant in the analysis of the charge transport in these materials.

3.2 Space-charge limited currents in thin PPV-

based hole-only devices

In order to study the effect of carrier accumulation at an Ohmic hole

contact hole-only diodes were fabricated from poly(2-methoxy,5-(2’-ethyl-hexoxy)-p-phenylene viny-lene) (MEH-PPV) with different layer thicknesses. Hole-only diodes from this MEH-PPV are prepared as follows: on top of a glass

substrate a transparent electrode, indium-tin oxide (ITO), has been patterned to form the hole injecting electrode. As a first step an anode of a hole-conducting

layer of poly(3,4-ethylenedioxythiophene)/ poly(styrenesulfonate) (PEDOT:PSS) is spincoated. Then on top of the PEDOT:PSS, MEH-PPV films ranging in thickness from 40 nm to 320 nm have been spin coated from toluene solution. The

device was finished by thermal evaporation of 100 nm of gold (Au) through a shadow mask. The hole-only diodes have been measured under controlled N2

atmosphere. The electrical measurements have been performed using a Keithley 2400 SourceMeter.

In Figure 3-2 the current density-voltage (J-V) measurements are presented

for MEH-PPV hole-only diodes with thicknesses L of 40 nm and 318 nm, respectively. The applied voltage is corrected for the built-in voltage Vbi of 0.4 V,

resulting from the work function difference between PEDOT:PSS and the evaporated Au electrode. Corrected with this built-in voltage the J-V curves show a quadratic behaviour at low (< 1 V) effective voltage (V-Vbi) for all thicknesses. For

thick devices a large voltage range is measured and the exact choice of Vbi is not critical. Also shown are fits with a SCLC model that assumes that at V=0 V there

are no charge carriers present in the device. In order to fit the J-V at higher voltages a density- and field dependent mobility has been taken into account [2].

For the device with a thickness of 318 nm a low density mobility µh(0,T) was found of 3.0×10-11 m2/Vs at room temperature, which is in agreement with earlier

Chapter 3

36

reported values [3]. However, for the 40 nm device a room temperature mobility was found of 8.7×10-10 m2/Vs. Apparently, the mobility seems to increase by more

than an order of magnitude when going from a thick (318 nm) to a thin (40 nm)

device. In order to study this mobility increase in more detail devices with a series

of different thicknesses have been measured and modeled. In Figure 3-3 the room temperature mobility µh as obtained from the conventional SCLC model is shown as a function of MEH-PPV layer thickness.

0 6 1210-4

10-3

10-2

10-1

100

101

J (A

/m2 )

V-Vbi (V)

T=292K T=275K T=255K T=235K T=215K

(a) 318nm

110-1

100

101

102

103

J (A

/m2 )

V-Vbi (V)

T=293K T=274K T=254K T=235K T=215K

40nm(b)

Figure 3-2. Temperature dependent J-V characteristics of MEH-PPV hole-

only diodes (empty symbols) with polymer thicknesses L of (a) 318nm and

(b) 40nm. The solid lines represent the fits with the SCL model without

taking into account diffusion of charge carriers.


37

Figure 3-3 shows that with decreasing layer thickness the mobility

gradually starts to increase, with especially a strong increase for thicknesses below 100 nm.

3.3 Modelling of the diffusion of holes from an

Ohmic contact at zero-bias into a thin film

In order to investigate whether this mobility increase originates from the

diffusion of charge carriers from the contacts into the device we have to calculate the charge carrier profiles at V=0. The amount of charge carriers in the devices will strongly depend on the boundary condition that is used for the injecting contact,

located at x=L. From earlier measurements on MEH-PPV it is known that the width of the Gaussian DOS typically amounts to 0.1-0.11 eV [3]. Furthermore, the

middle of the HOMO is located at around 5.35 eV below the vacuum level [4]. Together with the work function of PEDOT:PSS of 5.0 eV, as measured by a Kelvin probe, this gives a hole density of 5×1023 m-3 at the injecting

PEDOT:PSS/MEH-PPV interface, as shown in Figure 3-4. For the Au contact (x=0) a value of 1×1015 m-3 was used, that was derived from the increase of the

Figure 3-3. Distribution of the room temperature mobility µh(0,300 K)

as function of MEH-PPV layer thickness.

0 100 200 30010-11

10-10

10-9

(m2 /V

s)

L (nm)

Chapter 3

38

injection barrier with Vbi , but the calculations are not sensitive to the choice on this boundary.

It should be noted that the work function of PEDOT:PSS might be modified due to interface dipoles when brought in contact with organic

semiconductors like pentacene [6]. On the other hand, electrochemical modification of the PEDOT work function in PPV-based solar cells showed that the open-circuit voltage exactly followed the behaviour as expected from the

PEDOT work function [7]. This shows that the PEDOT work function is not pinned in this system, which supports our estimate of the hole density at the

interface. We use a recently developed drift-diffusion model to calculate the distribution of holes p(x) as a function of distance x in the device [8]. The injecting contact is taken at x=L. The inset of Figure 3-5 shows the hole distribution for the

device with a thickness of 40 nm. The hole density nearly exponentially decreases as a function from the injecting contact. In order to quantify the amount of charge

that flows into a device to equilibrate the Fermi-level we will use the average charge density ρav that is calculated from the distribution of p(x), given by

( )

L

dxxp

L

av

∫= 0ρ (3-1)

Figure 3-4. Boundary condition at the injecting PEDOT:PSS/MEH-PPV

contact (x=L). For the total number of sites in the Gaussian DOS with a

width σ=0.11 eV a value N of 3×1026 m

-3 was used [5].


39

In Figure 3-5 the increase of ρav is shown for decreasing device thickness.

It typically increases from 5×1021 m-3 for the 318 nm device to 2×1022 m-3 for the 40 nm device.

As a next step we calculate the J-V characteristics at room temperature,

combining the presence of charge carrier accumulation regions close to the contact with the density dependence of the hole mobility as given by Eq. 3-2.

1

3

0

4

0

0 0

0

)2(

sin

),0(),( −

+= TT

TT

c

hh pB

T

T

T

T

eTTp

α

πσ

µµ (3-2)

In this equation µh(0,T) is the hole mobility at low densities obtained from the quadratic SCL current, σ0 is a conductivity prefactor, α

-1 is the effective

0 100 200 300

5.0x1021

1.0x1022

1.5x1022

2.0x1022

0 10 20 30 401014

1019

1024

av (m

-3)

L (nm)

( m

-3)

x (nm)

40nm

Figure 3-5. Numerically calculated distribution of the average hole

density ρav as a function of polymer layer thicknesses L at zero bias

(V=0V).

Chapter 3

40

overlap parameter between localized states, T0 is a measure of the width of the exponential density of states, and Bc is the critical number of the onset of

percolation. From temperature dependent field-effect measurements the mobility

parameters T0, σ0, α-1 have been obtained; T0 = 540 K, σ0=3.1x107 S/m and α-1=

0.14 nm [5]. For Bc a value of 2.8 was used [9]. In Figure 3-6 the experimental J-V measurements for the MEH-PPV hole only diodes with different thicknesses are shown (symbols). The solid lines represent the calculated J-V characteristics taking

drift and diffusion and the density dependent mobility into account. All thicknesses can be described using the abovementioned parameters for the density dependence

of µh(p,T) and a µh(0,300 K) of 1.25×10-11m2/Vs .

This demonstrates that indeed the diffusion of charge carriers from the contacts into the polymer is responsible for the apparent increase of the charge

carrier mobility. Even for the thick device of 318 nm a low-density mobility of 3.0×11-11m2/Vs was obtained from the standard SCL model, whereas analysis of

the complete thickness dependence including diffusion leads to an intrinsic mobility of 1.25×10-11m2/Vs for MEH-PPV.

Figure 3-6. J-V characteristics of MEH-PPV hole-only diodes (empty

symbols) with different thicknesses L of the polymer. The solid lines

represent the calculations from a drift-diffusion SCL model that takes into

account the presence of holes at zero bias due to diffusion from the Ohmic

contact, combined with a density dependent mobility (Eq. 3-2).

0 2 4 6 8

10-3

10-1

101

103

J (A

/m2 )

V (V)

L=318nm L=152nm L=103nm L=56nm L=40nm


41

These results show that extrinsic carriers that diffuse in from Ohmic contacts clearly influence the results of charge transport studies in these materials.

A decreasing mobility with increasing layer thickness has also been

observed in transient measurements on PPV [10]. On short time scales the hole transport is dispersive, and for thin layers the holes do not relax to the deeper sites of the density of states giving rise to faster transit times. In that case, a systematic

decrease of the transient ‘mobility’ was observed, from layers of 100 nm down to 460 nm. Even for 460 nm layers this transient mobility was still a factor of 3 above

the measured DC mobility as obtained from J-V measurements. In our DC J-V measurements, where the contribution of fast and slow carriers are integrated in time, the mobility is already close to the intrinsic value at thicknesses >200 nm, as

shown in Figure 3-3. As a result we do not expect that relaxation effects play a large role in the interpretation of our data.

It has been demonstrated that the mobility-density relation can be directly obtained from the J-V characteristics [11]. The SCL current, with a density

dependent mobility according to Eq. (3-2) through a device with thickness L, could be accurately approximated by J=0.8epavµav(pav)Eav with Eav=V/L, pav is the average

density in the device given by )/(5.1 2

0 eLVp rav εε= [12], and µh(pav) the

mobility at density pav. Using an experimental J-V characteristic Eav and pav can be directly calculated for every applied voltage V and the corresponding J then directly gives the mobility µav using µav(pav) = J/(0.8epavEav). The resulting µav ~ pav

relation then showed a consistent correspondence with the field-effect mobilities from FET measurements [11]. However, in case of thin devices it appears that this

approximation has to be corrected for the average amount of carriers ρav0 that are already present in the device at V=0 V, given by

avavavavav EppeJ )()(8.0 00 ++= ρµρ (3-3).

For the different thicknesses the ρav0 is used from the numerical

calculations as shown in Figure 3-5. In Figure 3-7 the obtained µav( pav) relation is shown for devices with a layer thickness of 40 nm, 103 nm and 318 nm, together

with the mobilities obtained from field-effect measurements on MEH-PPV. Due to the high carrier concentrations in the thin devices the ‘gap’ between the hole-only diode data and the FET data has been closed, in contrast to earlier reported results

on thicker samples [11]. Figure 3-7 then shows the full experimentally determined mobility vs. density relation for MEH-PPV.

Chapter 3

42

3.4 Conclusions

In conclusion, we demonstrate that the mobility as obtained from space-charge

limited polymeric diodes is enhanced by diffusion of charge carriers from an

Ohmic contact into the device. For devices of only 40 nm thickness the mobility

value is enhanced by more than one order of magnitude. The apparent increase of

the mobility is described by a combination of a drift-diffusion model with a density

dependent mobility. Omission of diffusion, as is done in the standard SCLC model,

leads to an overestimation of the intrinsic charge carrier mobility in thin PLEDs. In

the next chapter the influence of the background carrier density that has diffused

from the Ohmic contact into the film on the temperature dependence of the

transport will be discussed.

Figure 3-7. Hole mobility as function of hole density in a hole-only diode

field effect transistor for MEH-PPV (symbols). The dashed line is

calculated using Eq. (3-1).

1019

1021

1023

1025

10-11

10-10

10-9

10-8

10-7

µ HO,

µ FE

T (

m2/V

s)

ρ (m-3)

FET

L=103nm

L= 40nm

L=318nm


43

REFERENCES

[1] J. G. Simmons, J. Phys. Chem. Solids 32, 1987 (1971).

[2] W. F. Pasveer, J. Cottaar, C. Tanase, R. Coehoorn, P. A. Bobbert, P. W. M. Blom, D. M. de Leeuw, and M. A. J. Michels, Phys. Rev. Lett. 94,

206601 (2005). [3] D. E. Markov, C. Tanase, P. W.M. Blom, J. Wildeman, Phys. Rev. B. 72,

045217 (2005).

[4] L. Bozano, S. A. Carter, J. C. Scott, G. G. Malliaras and P. J. Brock, Appl. Phys. Lett. 74, 1132 (1999)

[5] C. Tanase, E. J. Meijer, P. W. M. Blom, and D. M. de Leeuw, Phys. Rev. Lett. 91, 216601 (2003)

[6] N. Koch, A. Elschner, J. P. Rabe, and R. L. Johnson, Adv. Mater. 17, 330

(2005) [7] H. Frohne, S. E. Shaheen, C. J. Brabec, D. C. Muller, N. S. Sariciftci, and

K. Meerholz, Chem. Phys. Chem. 3, 795 (2002) [8] L. J. A. Koster, E. D. Smits, V. D. Mihailetchi, and P. W. M. Blom, Phys.

Rev. B. 72, 085205 (2005).

[9] G. E. Pike and C. H. Seager, Phys. Rev. B 10, 1421 (1974). [10] P. W. M. Blom and M. C. J. M. Vissenberg, Mater. Sci. and Eng. 27, 53

(2000) [11] C. Tanase, P. W. M. Blom, and D. M. de Leeuw, Phys. Rev. B.70, 193202

(2004)

[12] M. A. Lampert and P. Mark, Current Injection in Solids (Academic, New York, 1970).

Chapter 3

44

45

Chapter 4

Universal Arrhenius Temperature

Activated Charge Transport In

Diodes from Disordered Organic

Semiconductors

Abstract Charge tranport models developed for disordered organic semiconductors

predict a non-Arrhenius temperature dependence ln(µ) ∝1/T2 for the mobility µ.

We demonstrate that in space-charge limited diodes the hole mobility (µh) of a large variety of organic semiconductors shows an universal Arrhenius temperature

dependence µh(T)=µ0exp(-∆/kT) at low fields, due to the presence of extrinsic carriers from the Ohmic contact. The transport in a range of organic semiconductors, with a variation in room temperature mobility of more than six

orders of magnitude, is characterized by an universal mobility µ0 of 30-40 cm2/Vs.

As a result we can predict the full temperature dependence of their charge transport

properties with only the mobility at one temperature known.

Chapter 4

46

4.1 Introduction

The charge transport properties of conjugated polymers as poly(p-

phenylene vinylene) (PPV) and its derivatives have been extensively studied in order to understand the fundamental phenomena that govern the operation of these

devices. As explained in Chapter 1 the hole current in PLEDs at room temperature is space-charge limited (SCL) with a carrier density dependent mobility. The hole mobility is constant for charge carrier densities typically <1022 m-3 and increases

with a power law with density for carrier densities >1022 m-3. The occurrence of two regimes in the density dependence of the mobility is governed by the amount

of charge carriers in the Gaussian density of states (DOS), as schematically shown in Figure 4-1.

Figure 4-1. In Fig. 4-1a the hopping in a Gaussian DOS is schematically

represented for low carrier densities. The equilibrium level Eeq = -σ2/kT

represents the maximum of the density of occupied states (DOOS) and can

be regarded as the starting point for the hopping process towards the

transport level Et [1]. With increasing density, shown in Figure 4-1b, the

Fermi level Ef will at some point pass Eeq and serve as the new starting

point for hopping, leading to a density dependent mobility.

Universal Arrhenius Temperature Activated Charge Transport In Diodes…

47

For low carrier densities (Fig. 4-1a) the Fermi level Ef is below the

equilibrium level Eeq, which represents the maximum of the density of occupied states (DOOS), given by the product of the Gaussian DOS and the Fermi-Dirac

distribution. As a result Eeq can be regarded as the starting point of the hopping process towards the transport level Et, as indicated by the arrow. As long as Ef is below the equilibrium level Eeq the hopping probability and mobility is

independent on the position of Ef and thus independent on the carrier density. However, with increasing density (Fig. 4-1b) Ef will pass Eeq at a given density,

and then Ef serves as a new starting point for the hopping transport. With increasing density Ef moves closer to Et, leading to an enhanced mobility for larger densities. The relative positions of Ef and Eeq are crucial to explain the temperature

dependence of the charge transport [2,3]. Upon cooling down Eeq moves deeper into the tail of the Gaussian DOS, leading to a non-Arrhenius temperature

dependence given by ln (µ) ∝ 1/T2 in case of low densities [1]. For higher densities (Fig. 1b) the transport is governed by Ef and Et that are both only weakly

temperature dependent, leading to an Arrhenius like ln (µ) ∝ 1/T behaviour as often observed in organic transistors [4,5].

4.2. Temperature dependence of the hole mobility of

organic SCL diodes

The question whether the transport in organic semiconductors is better

described by a ln (µ) ∝ 1/T2 or ln (µ) ∝ 1/T (Eq. 1) behavior has been experimentally addressed by Borsenberger et al. [6] From time-of-flight (TOF) measurements on vapor-deposited 1,1-bis(di-4-tolylaminophenyl) cyclohexane

they convincingly demonstrated that the non-Arrhenius ln (µ) ∝ 1/T2 provides a more consistent description of the charge transport. However, in the analysis of various charge transport measurements an important aspect has been ignored: in

TOF studies thick samples (>1 µm) are used with blocking contacts to prevent charge injection. In SCL diodes on the other hand much thinner devices (50-300

nm) are measured, with at least one Ohmic contact to inject charge carriers. In order to align the Fermi level charge carriers diffuse from the Ohmic contact into the device. This results in the formation of an accumulation layer of free charge

carriers adjacent to the contact with a width of typically ten nanometer. Due to the density dependence of the mobility the hole transport in this accumulation region is

strongly enhanced, for devices of only 40 nm thickness even more than one order

Chapter 4

48

of magnitude [7], as shown in Chapter 3.. As a result, there is always a background of extrinsic charge carriers present in SCL diodes that modify the transport.

As a first step the temperature dependence of the hole mobility of SCL diodes made from poly(2-methoxy,5-(2’-ethyl-hexoxy)-p-phenylene viny-lene)

(MEH-PPV) with different layer thicknesses is investigated. The hole-only diodes were prepared on top of a glass substrate with a patterned transparent electrode, indium-tin oxide (ITO) on which an anode of a hole-conducting layer of poly(3,4-

ethylenedioxythiophene)/poly (styrenesulphonic acid) (PEDOT:PSS) is spincoated. Then on top of the PEDOT:PSS, MEH-PPV films ranging in thickness from 40

nm to 322 nm have been spin coated from toluene solution. The devices were finished by thermal evaporation of 100 nm of gold (Au) through a shadow mask. In

Fig. 4-2 the low field mobilities µh(E=0,T) of the samples with various thickness

are shown in an Arrhenius plot.

It is observed that the thinnest devices not only exhibit a higher mobility

[7], but also a weaker temperature activation. This is consistent with the fact that the thinnest devices have the largest concentration of background carriers that have diffused in from the Ohmic contact. As discussed in Chapter 3 the amount of

Figure 4-2. Measured low-field mobilities µh(E=0,T) for MEH-PPV based

SCL diodes with thicknesses ranging from 40 nm to 322 nm. The solid lines

are fits using the Arrhenius temperature dependence µh(E=0,T)=µ0exp(-

∆/kT).

0 2 4

10-14

10-11

10-8

10-5

10-2

µ(Ε=

0,Τ)

(m2/V

s)

1000/T (K-1)

40nm

56nm

89nm

103nm

156nm

248nm

318nm

322nm


49

charge that flows into a SCL diode in order to equilibrate the Fermi-level can be quantified by the average charge density ρav, given by

( )

L

dxxp

L

av

∫= 0ρ (4-1),

where p(x) is numerically calculated from a drift-diffusion program. The

simulations show that ρav typically increases from 5×1021 m-3 for the 318 nm device

to 2×1022 m-3 for the 40 nm device. Furthermore, the Arrhenius plot also shows that the temperature activated transport for the various MEH-PPV layer thicknesses all

originate from a single value of µ0=30 cm2/Vs. Such behaviour is expected since in

the limit T→∞ the charge transfer rates are only limited by the wave function overlap between neighbouring sites, and do not depend on layer thickness or filling

of the DOS. As a comparison the same data are also plotted in a ln(µ) ∝ 1/T2 plot,

as shown in Figure 4-3.

Figure 4-3. Measured low-field mobilities µh(E=0,T) for MEH-PPV based

SCL diodes with thicknesses ranging from 40 nm to 322 nm as a function

of ∝ 1/T2. The solid lines are fits using the mobility

relation

−= ∞

2

3

2exp

TkB

GDM

σµµ [1].

0.0 1.0x10-2

2.0x10-2

10-14

10-11

10-8

µ(Ε

=0,Τ

) (m

2/V

s)

1000/T2 (K

-1)

40mn

56nm

89nm

103nm

156nm

248nm

318nm

322nm

Chapter 4

50

Here, he µ0 varies more than order of magnitude, indicating that the

Arrhenius like ln (µ) ∝ 1/T behaviour is more consistent for these SCL diodes. Using a Gaussian DOS characterized by the total amount of sites N =3×1026 m-3

and σ = 0.11 eV [8] the equilibrium level Eeq is located at -0.46 eV at for T = 300 K. For a 40 nm diode with ρav = 2×10

22 m-3 the Fermi-level Ef is also located at -

0.46 eV. Upon cooling down Eeq sinks deeper into the tail of the DOS, whereas Ef is relatively temperature independent. As a result during a temperature scan from

200-300K Ef is always above Eeq, as indicated in Fig. 4-1b. Even for the 318 nm device with ρav = 5×10

21 m-3 , with Ef located at -0.51 eV, the equilibrium level Eeq will pass Ef already at T = 275 K, meaning that also for the thicker diodes Ef is

above Eeq for most of the temperature scan. This explains why an Arrhenius like behaviour is more applicable for SCL diodes that carry extrinsic background

carrier due to the presence of an Ohmic contact.

Since the observed µ0 for T→∞ is independent of the layer thickness of the SCL diodes it is interesting to compare the mobilities in a whole range of organic

semiconductors, as summarized in Table 1.

Polymer

name

Mobility

µ300K(m2/Vs)

and

Acrivation

energy (eV)

Chemical structure

PCBM 2×10-7 [9]

0.23 eV O

O

rr-P3HT 1.3×10-8 [10]

0.3 eV S

*S* n

rir-P3HT 2.8×10-9 [8]

0.35 eV

S* *n


51

DEH-PPV 1.2×10-9 [11] 0.359 eV

O

O

*

*n

MEH-PPV 5×10-11 [11]

0.45 eV O

O

*

*n

OC1C10-PPV 5×10-11 [12]

0.48 eV

O

O

*

*n

NRS-PPV 1.5×10-12

[10] 0.547 eV

O

O

O

*

*

n

PFO-TPD

1:7

3.7×10-13

[10] 0.59 eV

n

OO

NNO

OB

1NN Br

7

TABLE 1. Room temperature mobility µh(E=0,T=300K), activation energy ∆, and

structural formula of a number of organic semiconductors

The room temperature mobilities µh(E=0,T=300K) vary over six orders

from 10-13 cm2/Vs to 10-7 cm2/Vs. In Fig. 4-4 the mobilities, all obtained from SCL diodes, are shown in an Arrhenius plot. As expected, the materials with a higher mobility also show weaker temperature dependence due to the reduced disorder. In

Fig. 4-5 the Arrhenius plot is extrapolated to T→∞:

Chapter 4

52

Figure 4-4. Measured low-field mobilities µh(E=0,T) on SCL diodes for a

large range of disordered organic semiconductors (see Table 1). The solid

lines are fits using the Arrhenius temperature dependence

µh(E=0,T)=µ0exp(-∆/kT).

0 2 4 610

-19

10-16

10-13

10-10

10-7

10-4

10-1

PCBM

rir-P3HT

DEH-PPV

MEH-PPV

OC1C

10-PPV

NRS-PPV

rr-P3HT

PFO-TPD 1:7

µ(Ε=

0,Τ)

(m2/V

s)

1000/T (K-1)

µ0=30cm

2/Vs

Figure 4-5. Measured low-field mobilities µh(E=0,T) on SCL diodes for a

large range of disordered organic semiconductors (see Table 1). The solid

lines are fits using the Arrhenius temperature dependence

µh(E=0,T)=µ0exp(-∆/kT).

4 5

10-16

10-14

10-12

10-10

10-8

10-6

OC1C

10

µ(Ε

=0,Τ

) (m2/V

s)

1000/T (K-1)

rir-P3HT

DEH

PCBM

NRS

MEH

rr-P3HT

PFO-TPD 1:7


53

Remarkably, again all measured mobilities originate from nearly one point,

again µ0=30±10 cm2/Vs. The occurrence of a single value for µ0 for such a large

variety of organic semiconductors is surprising since its value is expected to be

governed by the amount of electronic overlap between the chain segments. This result would indicate that the packing and resulting wave function overlap would be very similar for the materials studied here. However, it should be noted that the

measured mobilities for the conjugated polymers are an average between both intrachain and interchain hopping processes. For intrachain hopping the hopping

distance and localization length are about equal, leading to a T→∞ mobility in a Gaussian DOS of [14]

σ

υµ phea 2

0 = (4-2)

with a the hopping distance, υph the phonon attempt frequency. Using typical

values of a = 1 nm, υph = 1014 s-1, and σ = 0.1 eV yields a µ0 of 10 cm

2/Vs, which is

in the right order of magnitude. However, with intrachain conjugation lengths of typically 6-7 nm and interchain packing distances of ~0.4 nm a coupling to

microscopic parameters as wave function overlap is very difficult to make.

The fact that the temperature activation is governed by one universal

prefactor µ0=30 cm2/Vs also enables us to predict the activation of charge transport

Figure 4-6. Activation energy ∆ as a function of the low-field room

temperature mobility µh(0,300K) as obtained from SCL diodes. The solid line is

the prediction from Eq. 4-3.

10-12

10-10

10-8

10-6

0.2

0.4

0.6

PCBM

rr-P3HT

rir-P3HTDEH-PPV

MEH-PPV

OC1C

10-PPV

NRS-PPV

PFO-TPD 1:7

Activa

tio

n e

ne

rgy (

eV

)

µ (0,300 Κ) (m2/Vs)

Chapter 4

54

with only the mobility at one temperature, for example µh(0,300K), known. For

Arrhenius-like transport the mobility is given by )/exp(),( 0 kTTEh ∆−= µµ . In

that case it can be rewritten as

⋅=∆

300K) (0,300 0

µµ

k (4-3)

In Figure 4-6 the measured low-field room temperature mobilities

µh(0,300K) are shown together with the measured activation energy ∆ for the various organic semiconductors. The solid line is the prediction made from Eq. (4-

3). The good agreement for semiconductors that differ by six orders of magnitude

in µh(0,300K) shows the validity of the Arrhenius like transport in all of these SCL diodes.

4.3 Experimental In conclusion, we demonstrate that the mobility of a variety of organic

semiconductors are consistent with an Arrhenius temperature dependent transport

with a universal prefactor µ0=30 cm2/Vs. The carriers diffusing into the diodes

from the Ohmic contact lift the Fermi level above the equilibrium level, resulting

in an Arrhenius like temperature dependence in stead of the characteristic

ln(µh)~1/T2 behaviour as observed for blocking contacts. As a result the

temperature dependence of the charge transport can be predicted when the mobility

at one single temperature is known. With the transport of holes known in very thin

layers (Chapter 3) and their temperature dependence characterized (Chapter 4) we

will now address the development of polyfluorene hole transport layers in Chapter

5.


55

REFERENCES: [1] H. Bässler, Phys. Status Solidi B 175, 15 (1993). [2] S.D. Baranovskii, H. Cordes, F. Hensel, and G. Leising, Phys. Rev. B. 62,

7934 (2000).

[3] R. Coehoorn, W. F. Pasveer, P. A. Bobbert, and M. A. J. Michels, Phys. Rev. B. 72, 155206 (2005).

[4] M. C. J. M. Vissenberg and M. Matters, Phys. Rev. B 57, 12964 (1998). [5] E. J. Meijer, C. Tanase, P. W. M. Blom, E. van Veenendaal, B.-H.

Huisman, D. M. de Leeuw, T. M. Klapwijk, Appl. Phys. Lett. 80, 3839

(2002). [6] P. M. Borsenberger, L. Pautmeier, R. Richert, and H. Bässler, J. Chem.

Phys. 94, 8276 (1991). [7] N. I. Craciun, and J. J. Brondijk and P. W. M. Blom, Phys. Rev. B 77,

035206 (2008) [8] C. Tanase, E. J. Meijer, P. W. M. Blom, and D. M. de Leeuw, Phys. Rev.

Lett. 91, 216601 (2003).

[9] V. D. Mihailetchi, J. K. J. van Duren, P. W. M. Blom, J. C. Hummelen, R. A. J. Janssen, J. M. Kroon, M. T. Rispens, W. J. H. Verhees, and M. M.

Wienk, Adv. Funct. Mater. 13, 43 (2003). [10] Measured in this study. [11] D. E. Markov, C. Tanase, P. W. M. Blom, and J. Wildeman, Phys. Rev. B

72, 045217 (2005). [12] P. W. M. Blom, M. J. M. de Jong, and M. G. van Munster, Phys. Rev. B

55, R656 (1997). [13] C. Tanase, P. W. M. Blom, D. M. de Leeuw and E. J. Meijer, Phys. Stat. Sol. 6, 1236 (2004).

[14] H. C. F. Martens, P. W. M. Blom, and H. F. M. Schoo, Phys. Rev. B. 61, 7489 (2000).

Chapter 4

56

57

Chapter 5

Substituted Polyfluorene Based

Hole Transport Layer With

Tunable Solubility

Abstract

We report on the synthesis and electrical characterization of polyfluorene-

triarylamine based hole transport layers (HTL). The solubility of the HTL can be

tuned by adjustment of the chemical structure, without loss of the charge transport

properties. Double-layer polymer light-emitting diodes are constructed with a HTL

that is not soluble in toluene at room temperature, combined with a poly(p-

phenylene vinylene) (PPV) derivative based light-emitting layer. The addition of

the HTL enhances the efficiency of the PLED with 10% at higher voltages.

Chapter 5

58

5.1 Introduction

Organic light-emitting diodes (OLEDs) are considered as promising

candidates ranging from lighting to full-colour emissive displays. Basically, an organic LED consists of a thin organic electroluminescent layer sandwiched between two electrodes. Organic LEDs are typically made from either small

molecules [1] or from conjugated polymers [2]. Small molecule based devices are deposited with vacuum techniques, whereas soluble conjugated polymers are

processed from solution. An advantage of evaporated small molecule based devices is that the active part consists of various layers with various functions, leading to highly efficient devices [3,4]. These layers are chosen to have properties

such as hole and electron transport, hole or electron blockage and high emission. The active part of present state-of-the-art polymeric light-emitting diodes (PLEDs)

consists of only a single layer. The use of only a single electro-optic layer has a large fundamental disadvantage: due to the reduced electron transport in conjugated polymers most of the light in a PLED is generated close to the metallic

cathode. This metallic cathode acts as a quenching site for the generated excitons, thereby strongly reducing the efficiency of the PLEDs. These fundamental

limitations can be circumvented by using devices consisting of a number of active layers. A major problem for polymer based multilayer devices is the solubility of the materials used; a multilayer can not be fabricated when a spin casted layer

dissolves in the solvent of the subsequent layer. As discussed in Chapter 2 in recent years a number of approaches have been

developed to realize solution processed multilayer PLEDs. Here we follow the approach to tune the solubility by chemical modification [5]; copolymers with

selective solubility can be achieved without loss of the enhanced charged transport properties. In this study it was shown that by shortening the (2’-ethylhexyloxy) side chains, from poly[2,5-bis(2’ethylhexyloxy)-1,4-phenylenevinylene] (BEH-

PPV), to butoxy side chains the polymer poly[2,5-bis(butoxy)-1,4-phenylenevinylene] (BB-PPV) was obtained, which is only soluble in chloroform

in very low concentrations. Consequently, by tuning the ratio of the BEH- and BB- monomers the solubility could be adjusted over the whole spectrum of solvents, thereby preserving the enhanced charge transport properties. Using a hole transport

layer (HTL) of a BEH-BB-PPV copolymer in 1:3 ratio was only soluble in chloroform, a bilayer PLED was constructed with NRS-PPV as emitting layer.

However, a disadvantage of these PPV-based bilayer PLEDs is that the HOMO and LUMO levels of the HTL and light-emitting polymer aligned, such that the electrons were not blocked at the interface.

Substituted Polyfluorene Based Hole Transport Layer With Tunable Solubility

59

5.2 Bilayer PLEDs with a polyfluorene based hole

transport layer In an improved double-layer device the HTL should have a larger band-gap than the light emitting polymer layer (LEP). At the same time, the HOMO levels of these two polymers have to align in order to efficiently inject the holes from the

HTL into the LEP, as schematically shown in Fig. 5-1. Furthermore, the chemical structure of the HTL has to be designed in such a way that when the LEP layer is

spin-coated on top it will not be dissolved [see Fig. 5-1].

Figure 5-1. Schematic representation of a dual layer PLED with a hole transport

layer (HTL) that blocks electrons and a light-emitting polymer (LEP). The dual

layer PLED is sandwiched between an Ohmic PEDOT:PSS contact and a Ba/Al

cathode.

As a starting material for the wide band gap HTL we started with the blue-

emitting PFO, which has a band gap of 3.2 EV [6]. However, due to its deep HOMO level of 5.8eV it is very difficult to efficiently inject holes into PFO [7]. In

order to lift the HOMO level PFO has been functionalized with triarylamine based units. These dioctylfluorene-triarylamine conjugated copolymers combine excellent hole transport properties of the triarylamines with the processability of

Chapter 5

60

conjugated polymers [6]. Furthermore, because of the decrease of the ionization potential, typically from 5.8 eV for PFO to around 5.0 eV for the copolymers, the

hole injection from standard anodes as PEDOT:PSS very efficient [7]. In this study we combined the 2,2'-(9,9-dioctyl-9H-fluorene-2,7-diyl)bis(4,4,5,5-tetramethyl-

1,3,2-dioxaborolane) monomer (Figure 5-2 (1)) with N,N-di(4-ethylhexylphenyl)-N,N-bis(4-bromophenyl)-[1,1’-biphenyl]-4,4’-diamine (Figure 5-2 (2)), resulting in poly(9,9-dioctyl-9H-fluorene-2,7-diyl)co-(N4,N4'-bis(p-phenylene)-N4,N4'-

bis(4-(2-ethylhexyloxy)phenyl)biphenyl-4,4'-diamine) (PFO-BEHTPD) (Figure 5-2 (3)).

O

OB

O

OB B r

OO

NN B r

OO

NNn

+

P d c a t .

( 3 )

( 1 )

( 2 )

Figure 5-2. Chemical structure of PFO-BEHTPD (3). (1) 2,2'-(9,9-dioctyl-9H-

fluorene-2,7-diyl)bis(4,4,5,5-tetramethyl-1,3,2-dioxaborolane). (2) N,N-di(4-

ethylhexylphenyl)-N,N-bis(4-bromophenyl)-[1,1’-biphenyl]-4,4’-diamine.

As a next step the solubility of the PFO-BEHTPD, which is soluble in a

large amount of solvents, has to be tuned in order to make it suited as a HTL in a solution processed PLED. For this purpose the 2,2'-(9,9-dioctyl-9H-fluorene-2,7-diyl)bis(4,4,5,5-tetramethyl-1,3,2-dioxaborolane) (Figure 5-3 (1)) was combined

with N,N-diphenyl-N,N-bis(4-bromophenyl)-[1,1’-biphenyl]-4,4’-diamine (Figure 5-3 (2)) leading to poly(9,9-dioctyl-9H-fluorene-2,7-diyl)co-(N4,N4'-bis(p-

phenylene)-N4,N4'-diphenylbiphenyl-4,4'-diamine) (PFO-TPD) (Figure 5-3 (3)), which is an insoluble polymer due to the absence of sidegroups on the phenyl units. A tunable solubility can be achieved by copolymerization of these two

polymers, resulted in the random copolymer Poly[(9,9-dioctyl-9H-fluorene-2,7-diyl)co-(N4,N4'-bis(p-phenylene)-N4,N4'-bis(4-(2-

ethylhexyloxy)phenyl)biphenyl-4,4'-diamine)]x ran-[(9,9-dioctyl-9H-fluorene-2,7-diyl)co-(N4,N4'-bis(p-phenylene)-N4,N4'-diphenylbiphenyl-4,4'-diamine)]y ([PFO-


61

BEHTPD]x[PFO-TPD]y) (Figure 5-4 (4)). For a ratio of 1:7, ([PFO-BEHTPD]x[PFO-TPD]y) become insoluble in toluene at room temperature, but it is

still soluble when the solution is heated at temperatures above 70°C.

O

OB

O

OB NN BrBr

NNn

+

Pd cat.

(3)

(1) (2)

Figure 5-3. Chemical structure of PFO-TPD (3). (1) 2,2'-(9,9-dioctyl-9H-fluorene-

2,7-diyl)bis(4,4,5,5-tetramethyl-1,3,2-dioxaborolane). (2) N,N-diphenyl-N,N-bis(4-

bromophenyl)-[1,1’-biphenyl]-4,4’-diamine

Br

OO

NN BrO

OB

O

OB NN BrBr

NN *

OO

NN*1 7

n

+Pd cat.

(4)

(3)

+

(1) (2)

Figure 5-4. Chemical structure of [PFO-BEHTPD]1[PFO-TPD]7 (4). (1) 2,2'-

(9,9-dioctyl-9H-fluorene-2,7-diyl)bis(4,4,5,5-tetramethyl-1,3,2-dioxaborolane). (2)

N,N-diphenyl-N,N-bis(4-bromophenyl)-[1,1’-biphenyl]-4,4’-diamine. (3) N,N-

di(4-ethylhexyloxyphenyl)-N,N-bis(4-bromophenyl)-[1,1’-biphenyl]-4,4’-diamine

Chapter 5

62

As a next step we characterized the hole transport properties of the

dioctylfluorene-triarylamine copolymers. Hole-only diodes were fabricated from PFO-BEHTPD and [PFO-BEHTPD]1[PFO-TPD]7 using patterned indium-tin

oxide (ITO) on top of a glass substrate, followed by spincoating of a hole-injection layer of (PEDOT:PSS). Then PFO-BEHTPD or [PFO-BEHTPD]1[PFO-TPD]7 films have been spin coated from a hot toluene (70 °C) solution. The devices were

finished by thermal evaporation of 100 nm of gold (Au) through a shadow mask. The hole-only diodes have been measured under controlled N2 atmosphere using a

Keithley 2400 SourceMeter.

First, the hole mobility of the well souble PFO-BEHTPD is investigated. From earlier studies on derivatives of poly(p-phenylene vinylene) (PPV) it has

Figure 5-5. Temperature dependent J-V characteristics of a PFO-BEHTPD

hole-only diode (symbols) with polymer thicknesses L=300nm. The solid lines

represent the calculations from a drift-diffusion transport model that takes into

account the presence of holes at zero bias due to diffusion from the Ohmic

contact, combined with a density dependent hole mobility. The inset shows the

zero-field mobility as a function of temperature

0 2 4 6 8 10

10-3

10-2

10-1

100

101

102

3.5 4.0 4.5

10-12

10-11

10-10

µ(E

=0

,T)(

m2/V

s)

1000/T (K-1

)

J (

A/m

2)

V-Vbi (V)


63

been demonstrated that the hole transport is space-charge limited (SCLC) with a hole mobility µh of 5×10

−11 m2 /V s at low voltages at room temperature [8]. As

discussed in Chapter 1 the enhancement of the SCL current at higher voltages results from the dependence of the hole mobility on the charge carrier density. By a

combined study on field-effect transistors and polymeric diodes it was demonstrated that the hole mobility is constant for charge carrier densities <10 22 m−3 and increases with a power law for densities >1022 m−3 [9]. These experiments

revealed that for a complete description of the charge transport in conjugated polymers both the effects of carrier density and electric field on the mobility has to

be taken into account [10]. Furthermore, as explained in Chapter 3 the mobility obtained from SCL polymeric diodes is enhanced by diffusion of charge carriers from an Ohmic contact into the device. In Fig. 2 the experimental J-V

measurements for the PFO-BEHTPD hole only diodes with a thickness of 300nm are shown (symbols). The applied voltage is corrected for the built-in voltage Vbi.

The solid lines represent the calculated J-V characteristics using a device model taking drift, diffusion and the density dependent mobility into account [11]. At

room temperature from the analysis of the J-V characteristics we find a hole mobility at low electric fields of 1.2×10-10m2/Vs, which is two times higher as compared to the mobilities reported for PPV. The temperature dependence of the

measured low-field mobility µh(E=0,T) of PFO-BEHTPD is shown in the inset of figure 5-5; it follows an Arrhenius-like temperature dependence µh(E=0,T)=µ0exp(-

∆/kT), with an activation energy of 0.44 eV. The mobility µ0 extrapolated to T→∞

amounts to µ0=30±10 cm2/Vs, showing that the transport in PFO-BEHTPD also

follows the universal behavior between activation energy and mobility as shown in Chapter 4 for a whole range of conjugated polymers. Subsequently, the hole

mobility of [PFO-BEHTPD]1[PFO-TPD]7 is investigated. An important question is whether the modification of the solubility is of influence on the charge transport

properties. In Figure 5-6 the experimental J-V measurements for the [PFO-

BEHTPD]1[PFO-TPD]7 hole only diodes with a thickness of 420 nm are shown (symbols). From the analysis of the J-V characteristics a zero-field hole mobility of 1.0×10-10m2/Vs has been found, which is very close to the mobility of PFO-

BEHTPD. This shows that the mobility of the polymer does not change by the copolymerization with the unsoluble PFO-TPD. Furthermore, we verified that the

temperature dependence of the mobility was also indentical. In order to make a bi-layer device, a light-emitting polymer (LEP) has to

be spin-coated on top of the of the [PFO-BEHTPD]1[PFO-TPD]7 HTL at room

temperature. We choose as a light emitting polymer-methoxy, 5- (2’ ethyl-hexyloxy)-p-phenylene vinylene) (MEH-PPV), since it is a well characterized

polymer of which the charge transport properties are well known (chapter 3).

Chapter 5

64

Furthermore, with an average HOMO level of 5.1eV and a LUMO level of

2.7eV [12] MEH-PPV is a suitable candidate for the second polymer layer, because it is expected that holes can be easily injected from the [PFO-BEHTPD]1[PFO-

TPD]7 into the MEH-PPV, since there is no energy offset between the HOMO levels. The MEH-PPV was dissolved in toluene and spin-coated at room temperature. As a reference, in Figure 5-7 the J-V characteristics of a hole-only

device consisting of a single layer MEH-PPV of 90nm at room temperature are shown, together with the a double layer [PFO-BEHTPD]1[PFO-TPD]7/MEH-PPV.

The thickness of the polymer layers in the double layer are 50nm for [PFO-BEHTPD]1[PFO-TPD]7 and 85nm for MEH-PPV. The MEH-PPV in both single- and double layer devices are spincoated from the same solution so that a direct

comparison between the devices can be made. From the modeling of the single layer MEH-PPV device we obtained a zero-field room temperature of 3.0×10-

11m2/Vs. With now the mobility of MEH-PPV and [PFO-BEHTPD]1[PFO-TPD]7 known we can predict the J-V characteristics of the double layer device. The

predicted J-V characteristic is in very good agreement with the experimental data, as shown in Figure 5-7. This demonstrates that the thickness of the [PFO-

Figure 5-6. Temperature dependent J-V characteristics of [PFO-

BEHTPD]1[PFO-TPD]7 based hole-only diodes (symbols) with a polymer

thickness of L=420nm. The solid lines represent the calculations from the drift-

diffusion model.

0 5 10 15 20 25 30

10-3

10-2

10-1

100

101

102

103

J (

A/m

2)

V-Vbi (V)


65

BEHTPD]1[PFO-TPD]7 is not affected by spincoating MEH-PPV on top.

Finally, a PEDOT/[PFO-BEHTPD]1[PFO-TPD]7/MEH-PPV/Ba/Al dual

layer PLED was fabricated. The thickness of the polymer layers are 30nm for the [PFO-BEHTPD]1[PFO-TPD]7 and 85nm for the MEH-PPV, as obtained from thickness measurements using a DEKTAK profilometer. We applied only a thin

[PFO-BEHTPD]1[PFO-TPD]7 HTL in order to limit the voltage drop across the layer. In order to verify that there was still a HTL present in the dual layer device

we also performed optical absorption measurements, as shown in Figure 5-8. The absorption feature of the [PFO-BEHTPD]1[PFO-TPD]7 at 350 nm, which is not present in the MEH-PPV, is still clearly visible in the double-layer device. This

demonstrates that deposition of the MEH-PPV layer does not dissolve the HTL when spincoated at room temperature.

Because of the alignment of the HOMO levels of MEH-PPV and [PFO-BEHTPD]1[PFO-TPD]7 and the relatively large band gap of [PFO-

BEHTPD]1[PFO-TPD]7, as schematically indicated in Figure 5-1, the electrons injected from the cathode into the MEH-PPV will be blocked at the interface between the polymers. As a result in the double-layer PLED all the light will be

generated in the MEH-PPV layer.

Figure 5-7.J-V characteristics of a single layer MEH-PPV and dual layer [PFO-

BEHTPD]1[PFO-TPD]7/MEH-PPV hole-only diodes (symbols) at room

temperature. The solid lines are the model calculations from a drift-diffusion

model.

0 1 2 3 4 5

10-3

10-2

10-1

100

101

102

103

Double layer L=135nm

Single layer L=90nm

J (

A/m

2)

V-Vbi (V)

Chapter 5

66

Figure 5-9. J-V characteristics of a standard PEDOT:PSS/MEH-PPV/Ba/Al

PLED together with a PEDOT:PSS/ [PFO-BEHTPD]1[PFO-TPD]7/MEH-

PPV/Ba/Al dual-layer PLED at room temperature. The thickness of the [PFO-

BEHTPD]1[PFO-TPD]7 layer amounts to 30nm nm, the thickness of the MEH-

PPV layer is 85 nm in both devices.

0 1 2 3 4 5

100

101

102

103

J (

A/m

2)

V-Vbi (V)

MEH-PPV L=85nm

Double layer L=110nm

Figure 5-8. Optical density over wavelength for MEH-PPV (squares), PFO-

BEHTPD]1[PFO-TPD]7 (circles) and for the double layer (triangle).

300 400 500 600 700 800

0.0

0.5

1.0

1.5

optical density (

a.u

.)

wavelength (nm)


67

In Figure 5-9 the J-V characteristics of the double layer PLED is shown, together with the single layer MEH-PPV based PLED. In order to enable a direct

comparison we spincoated the MEH-PPV layer from the same solution. The J-V characteristic of the double layer PLED is slightly lower than the MEH-PPV single

layer device, due to the extra voltage drop across the HTL. This voltage drop not only depends on the mobility and thickness of the [PFO-BEHTPD]1[PFO-TPD]7 hole transport layer, but also on the fact that the current in the HTL is space-charge

limited. This means that in order to fill the HTL with charge carriers and make it conductive a certain voltage is required to electrostatically allow this space-charge.

In Fig. 5-10 the efficiency (light output / current) is plotted for a

PEDOT:PSS/MEH-PPV/Ba/Al and the double layer PEDOT:PSS/ [PFO-BEHTPD]1[PFO-TPD]7/MEH-PPV/Ba/Al. The efficiencies are normalized to the

maximum efficiency of the MEH-PPV based PLED. We observe that at lower voltages the efficiency of the double-layer PLED rises more slowly with voltage as

compared to the single layer PLED. The increase of the efficiency with voltage is a direct consequence of the unbalanced charge transport in the PLEDs [8]. The electron transport is strongly reduced as compared to the hole transport due to

trapping effects. As a result at low voltages the light is mainly generated in a region close to the cathode, whereas at high voltages, when the electron traps are

filled, the light is generated more homogeneously in the device. Due to the non-radiative recombination losses at the cathode interface, where a large number of exciton quenchers are present, the efficiency is strongly reduced at low voltages.

Figure 5-10. Normalized efficiency (light output/current) of a single layer MEH-

based and double layer [PFO-BEHTPD]1[PFO-TPD]7/MEH-PPV based PLED.

0 1 2 3 4 5 6

0.2

0.4

0.6

0.8

1.0

CE

/CE

max

V-Vbi (V)

Double layer LED

Single layer LED

Chapter 5

68

The slower rise of the efficiency for the double-layer device is also a consequence

of the additional voltage drop across the HTL. In order to fill the electron traps and

reduce the quenching a higher total voltage is required. At higher voltages, where

the light is generated more homogeneously in the MEH-PPV layer, the efficiency

of the double layer device is 10% higher as compared to the single layer PLED.

The presence of the HTL with it electron blocking functionality reduces the

quenching of excitons at the PEDOT:PSS anode, thereby enhancing the efficiency.

5.3 Conclusions

In conclusion, we have reported on the synthesis and electrical characterization of polyfluorene-triarylamine copolymers with a tunable solubility. The room

temperature mobility amounts to ~1×10-10m2/Vs and is not affected by the addition of non-soluble derivatives. The [PFO-BEHTPD]1[PFO-TPD]7 HTL is insoluble in

toluene at room temperature, enabling the construction of dual layer devices with MEH-PPV as active layer. These double layer PLEDs exhibit a 10% higher efficiency at higher voltage due to a reduced exciton quenching at the PEDOT:PSS

anode. For a further increase of the efficiency an additional electron transport/hole blocking layer will be required. However, the analysis of the electron transport in

conjugated polymers is strongly hindered by the presence of hysteresis in the J-V characteristics. In the next chapter the origin of this hysteresis will be addressed.


69

REFERENCES: [1] C.W. Tang and S.A. VanSlyke, Appl. Phys. Lett. 51, 913 (1987). [2] J. H. Burroughes, D. D. C. Bradley, A. R. Brown, R. N. Marks, K. Mackey, R.

H. Friend, P. L. Burn, and A. B. Holmes, Nature 347, 539 (1990). [3] B.W. D’Andrade, M.A. Baldo, C. Adachi, J. Brooks, M.E. Thompson, and S.R. Forrest, Appl. Phys. Lett. 79, 1045 (2001).

[4] S. Reineke et al., Nature 459, 234 (2009). [5] C. Tanase, J. Wildeman and P.W.M. Blom, Adv. Funct. Mater. 15, 2011

(2005). [6] M. Redecker, D.D.C. Bradley, M. Inbasekaran, E.P. Woo, Appl. Phys. Lett. 73, 1565 (1998).

[7] A. J. Campbell, D. D. C. Bradley, H. Antoniadis, J. Appl. Phys. 89, 3343 (2001).

[8] P. W. M. Blom, M. C. J. M. Vissenberg, Mat. Sc. and Engineering 27, 53 (2000).

[9] C. Tanase, E. J. Meijer, P. W. M. Blom, and D. M. de Leeuw, Phys. Rev. Lett. 91, 216601 (2003). [10] W. F. Pasveer, J. Cottaar, C. Tanase, R. Coehoorn, P. A. Bobbert, P. W. M.

Blom, D. M. de Leeuw, and M. A. J. Michels, Phys. Rev. Lett. 94, 206601 (2005). [11] L. J. A. Koster, E. C. P. Smits, V. D. Mihailetchi, and P. W. M. Blom, Phys.

Rev. B. 72, 085205 (2005). [12] A. L. Holt, J. M. Leger, and S. A. Carter, J. Chem. Phys. 123, 044704 (2005).

Chapter 5

70

APPENDIX:

Synthesis of Poly(9,9-dioctyl-9H-fluorene-2,7-diyl)co-(N4,N4'-bis(p-

phenylene)-N4,N4'-bis(4-(2-ethylhexyloxy)phenyl)biphenyl-4,4'-diamine)

(PFO-BEHTPD). Under a nitrogen atmosphere, 216 mg (0.34 mmol) of 2,2'-(9,9-

dioctyl-9H-fluorene-2,7-diyl)bis(4,4,5,5-tetramethyl-1,3,2-dioxaborolane) (Figure 5-2 (1)), 307 mg, 0.34 mmol) of N,N-di(4-ethylhexylphenyl)-N,N-bis(4-

bromophenyl)-[1,1’-biphenyl]-4,4’-diamine (Figure 5-2 (2)), 8 mg (.007mmol) of tetrakis(triphenylphoshine)palladium (0), toluene (10 ml), KOH solution (4mL, 20 %) and 13.5 mg of TBABr (ca 0.04 mmol) were placed in a round bottom flask and

stirred vigorously at refluxed temp for 4 h. The reaction mixture was worked up by precipitating it in 40 ml of methanol and stirred for 0.5 hour. The solid green-

yellow material was washed with methanol and aceton and isolated on a buchner filter. The polymer is dried in air. Yield 350 mg, 93 %. The polymer is further

purified by dissolving it in 4 mL of toluene and precipitating in aceton (50 mL), affording 335 mg of pure PFO-BEHTPD (Figure 5-2 (3)).

Synthesis of Poly(9,9-dioctyl-9H-fluorene-2,7-diyl)co-(N4,N4'-bis(p-

phenylene)-N4,N4'-diphenylbiphenyl-4,4'-diamine) (PFO-TPD). Under a nitrogen atmosphere, 321.3 mg (0.5 mmol) of 2,2'-(9,9-dioctyl-9H-fluorene-2,7-

diyl)bis(4,4,5,5-tetramethyl-1,3,2-dioxaborolane) (Figure 5-3 (1)), 323.2 mg (0.5 mmol) N,N-diphenyl-N,N-bis(4-bromophenyl)-[1,1’-biphenyl]-4,4’-diamine

(Figure 5-3 (2)), 9 mg (.008mmol) of tetrakis(triphenylphoshine)palladium (0), toluene (10 ml), KOH solution (5mL, 20 %) and 17 mg of TBABr (ca 0.05 mmol) were placed in a round bottom flask and stirred vigorously at refluxed temp for 4 h,

(after 1 h already a lump of sticky material was observed). The reaction mixture was diluted with ca 30-40 mL chloroform (gel-like material), precipitated in 400

ml of methanol and stirred for 0.5 hour. The formed precipitate was washed with methanol and isolated on a Buchner filter. The polymer is dried, affording 418 mg, 95 % of crude polymer (Figure 5-3 (3)). The crude polymer was redissolved in hot

dichlorobenzene (40 mL) and reprecipitated from acetone. The refined polymer was isolated on a filter and dried in vacuum at room temperature. Yield 780 mg, 90

%. The molecular weight was determined with NMR: 20.000 g/mol

Synthesis of Poly[(9,9-dioctyl-9H-fluorene-2,7-diyl)co-(N4,N4'-bis(p-

phenylene)-N4,N4'-bis(4-(2-ethylhexyloxy)phenyl)biphenyl-4,4'-diamine)]1


71

ran-[(9,9-dioctyl-9H-fluorene-2,7-diyl)co-(N4,N4'-bis(p-phenylene)-N4,N4'-

diphenylbiphenyl-4,4'-diamine)]7 [PFO-BEHTPD]1[PFO-TPD]7. Under a

nitrogen atmosphere, 282 mg (0.44 mmol) of 2,2'-(9,9-dioctyl-9H-fluorene-2,7-diyl)bis(4,4,5,5-tetramethyl-1,3,2-dioxaborolane) (Figure 5-4 (1)), 248,5 mg (0.385

mmol) N,N-diphenyl-N,N-bis(4-bromophenyl)-[1,1’-biphenyl]-4,4’-diamine (Figure 5-4 (2)) 45.5 mg (0.055 mmol) N,N-di(4-ethylhexyloxyphenyl)-N,N-bis(4-bromophenyl)-[1,1’-biphenyl]-4,4’-diamine (Figure 5-4 (3)), 10 mg (.009 mmol) of

tetrakis(triphenylphoshine)palladium (0), toluene (12 ml), KOH solution (5mL, 20 %) and 16 mg of TBABr (ca 0.05 mmol) were placed in a round bottom flask and

stirred vigorously at refluxed temp for 3 h.. The reaction mixture was worked up by precipitating it in 40 ml of methanol and stirred for 0.5 hour. The solid material was washed with methanol and aceton and isolated on a buchner filter. The

polymer is dried in air. Yield 385 mg, 90%. The polymer was further purified by dissolving it in 15 ml hot toluene, and precipitating in aceton. (75 mL), affording

345 mg of pure random copolymer (Figure 5-4 (4)).

Device Fabrication. In order to characterize the hole transport hole-only diodes were fabricated from both PFO-BEHTPD and [PFO-BEHTPD]1[PFO-TPD]7. These hole-only diodes are prepared as follows: on top of a glass substrate a

transparent electrode, indium-tin oxide (ITO), has been patterned to form the hole injecting electrode. Subsequently an anode of the hole-conducting polymer

PEDOT:PSS is spincoated. Then on top of the PEDOT:PSS, PFO-BEHTPD or [PFO-BEHTPD]1[PFO-TPD]7 films have been spin coated from hot toluene (70°C) solutions. The devices were finished by thermal evaporation of 100 nm of

gold (Au) through a shadow mask. Au has a high work function that does not inject electrons into the polymer. The hole-only diodes have been measured under

controlled N2 atmosphere. The electrical measurements have been performed using a Keithley 2400 SourceMeter. It should be mentioned that [PFO-BEHTPD]1[PFO-TPD]7 is soluble only in hot toluene and it becomes insoluble in toluene at room

temperature. For the double layer PLEDs MEH-PPV was spincoated on top of the PFO-based HTL as light-emitting layer from toluene at room temperature.

Chapter 5

72

73

Chapter 6

Hysteresis-Free Electron Currents

In Poly(P-Phenylene Vinylene)

Derivatives

Abstract

The interpretation of electron currents in conjugated polymers is strongly

hindered by the occurrence of hysteresis. We investigate the transport of electrons

in electron-only devices based on derivatives of poly(p-phenylene vinylene) (PPV)

for various hole-blocking bottom electrodes as well as purification of the polymer.

The use of a variety of hole blocking bottom contacts, as metallic electrodes and n-

type doped polymers, did not give any improvement in the observed hysteresis. By

purification of the PPV, hysteresis free electron-only currents can be obtained. The

deep traps responsible for hysteresis, with a concentration in the 1016 cm-3 range,

are not responsible for the trap-limited electron transport as observed in purified

PPV.

Chapter 6

74

6.1 Introduction

Charge transport is an important issue with regard to the understanding and optimization of electronic devices made from conjugated polymers. In the last two

decades a large effort has been put on the characterization of the transport of holes, which is the dominant charge carrier. A major problem with the investigation of the electron transport is the construction of so-called electron only devices, where

hole blocking electrodes are required that are usually reactive. The resulting current density vs. voltage (J-V) characteristics of these devices often exhibit, next

to low currents, strong hysteresis effects. Hysteresis effects often originate from the presence of deeply trapped charges of which the escape time is longer than the time it takes to make the J-V sweep. As a result after the first scan the device is still

charged and out of equilibrium, and a subsequent J-V scan is influenced by the presence of these deeply trapped charges. This strongly hinders the interpretation

of the electron currents, especially in cases where a series of subsequent sweeps are made as, for example, a temperature scan. The presence and origin of the hysteresis in most of the electron currents has not been addressed so far. Major candidates

responsible for the strong hysteresis are electrons deeply trapped either in the bulk of the polymer or at the hole blocking electrode/ polymer interface. In this chapter

the transport of electrons in electron-only devices based on MDMO-PPV and poly[2-methoxy-5-(2-ethylhexyloxy)-1,4-phenylenevinylene] (MEH-PPV) for a variety of hole-blocking bottom electrodes, as well as purification of the polymer is

investigated.

6.2 Hysteresis in the electron currents of PPV-based

conjugated polymers

A major experimental problem with the investigation of the electron

transport is the construction of the electron only devices, where hole blocking bottom electrodes are required that are usually reactive. During spincoating of the polymer such a reactive bottom electrode might react with the polymer solution.

As a result next to deep traps in the polymer layer itself also electrons trapped at the hole blocking electrode/ polymer interface might be a source of hysteresis. In

order to discriminate between bulk trapping and interface trapping we fabricated

Hysteresis-Free Electron Currents In Poly(P-Phenylene vinylene) Derivatives

75

electron-only devices on a variety of metallic bottom electrodes. Next to the standard Al electrode also bottom electrodes of Yb, Ga, Sn, and In were used. All the electrodes were prepared by thermal evaporation at low pressure 10-7–10-6

mbar, with a thickness of 20 nm on top of indium tin oxide (ITO) coated glass. Subsequently either MDMO-PPV or MEH-PPV was spin coated from toluene

solution. For the synthesis of the MEH-PPV 500 mg of Gilch MEH monomer was dissolved in dry dioxane (0.02M). The mixture was stirred at 25 °C under a continuous flow of nitrogen. 2.6 equivalents of a KtBuO solution (0.87M in

dioxane) were added dropwise over a time period of 15 minutes to the stirred monomer solution. After a waiting period of 10 minutes another 2 equivalents of a

KtBuO solution (0.90M in dioxane) were added in one go. During the addition of base an insoluble gel is formed. The reaction proceeded for 2 hours at 25 °C under a nitrogen atmosphere. Subsequently, the temperature was set at 100 °C and the

mixture reacted for another 16 hours under dark conditions to decrease the amount of gel. After reaction the mixture was cooled and subsequently precipitated in 500

ml stirred cold methanol. The mixture was filtered and the polymer was collected. The total yield of the reaction amounts to 50%. In order to finish the electron –only

devices barium (Ba) top electrodes were vapor deposited and coated with a protective aluminum layer. All top electrode Current density-voltage J-V measurements were performed in the dark and under a N2 atmosphere, using a

computer-controlled source meter unit Keithley 2400. In Figure 6-1 subsequent J-V sweeps are shown for an Al/MEH-PPV/Ba/Al

electron- only device. The thickness of the MEH-PPV amounts to 140 nm. In the first sweep the voltage is scanned from 0-3 V and back, in the second sweep from 0-5 V and back, and at every subsequent scan the maximum applied voltage is

increased with 2V. In the first sweep the up-scan shows an electron current that is typical for MEH-

PPV, whereas the back-scan shows a strong decrease directly when going down from the maximum applied voltage on, and at a finite voltage VT1 (1.4 V) the electron current even decreases to values below the sensitivity of the set-up (J =10-

6 A/m2 as indicated by the dashed line in Figure 6-1, determined by the sensitivity of the source-measure unit). The up-scan of the second sweep then closely follows

the back-scan of the first sweep; first there is no measurable current and from VT1 it strongly increases. For voltages higher than the maximum voltage VMAX1of the first sweep (5V), the up-scan connects to the up-scan of the first sweep. The back-scan

of the second sweep then decreases again and becomes undetectable at a voltage VT2 (2.2 V). This behavior then repeats for every subsequent sweep. It should also

be noted that when a fresh device is directly scanned to a higher voltage of 9 V (Fig. 6-1, solid line), its up-scan forms the envelope of the up-scans of the other

Chapter 6

76

sweeps that were carried out with a lower maximum voltage. Identical results were also observed for electron-only devices based on MDMO-PPV.

This behavior can be explained as follows, schematically indicated in Figure 6-2.

We consider a semiconductor that is sandwiched between two Ohmic contacts. The semiconducting material contains both shallow- (s) and deep (d) traps. At zero bias the fresh device is free of trapped charges (2a). On application of a voltage the

device is charged with carriers from the contact, resulting in a trap-limited current. The total amount of carriers injected into the device at a given voltage V is

approximately given by C0V, with C0 the geometrical capacitance. For a trap-free space-charge limited device the amount of charges equals exactly 1.5 C0V, for a trap-limited device it is closer to C0V since the traps confine the carriers closer to

the injecting electrode [1]. Thus, at the maximum voltage VMAX1 the total amount of charges in the device typically amounts to C0VMAX.. These charges can either be

free Qf, trapped in shallow traps Qs that are in thermal equilibrium with the free carriers, and trapped in deep traps Qd from which they can not escape (6-2b). On

Figure 6-1 Subsequent J-V characteristic of a Al/MEH-PPV/Ba/Al electron-only

device with a thickness of 140 nm. With every sweep the maximum voltage is

increased with 2V.

0 1 2 3 4 5 6 7 8 9 10

10-610-510-410-310-210-1100101

VT4

VMAX4

VMAX3

VMAX2

J (A

/m2 )

V (V)

1st scan 2nd scan 3rd scan 4th scan 1st scan

VT1

VMAX1


77

the back-scan, when the voltage is lowered also the amount of charges C0V that is electrostatically allowed in the device will be lowered.

However, since Qd can not be removed only Qf and Qs will be decreased

(6-2c), leading to a very strong reduction of the current during the back-scan. As a result clockwise hysteresis will occur. At the voltage VT1 given by Qd = C0 VT1 the

charges Qf and Qs will be completely removed from the sample and the current will go to zero (6-2d), since the presence of free charges is electrostatically not allowed anymore. For lower voltages than VT1 (1.4 V) the current will remain zero, since Qd

exceeds C0V and the system is completely out of equilibrium. For the second sweep no current will flow until the applied voltage reaches VT1. For V>VT1 it is

electrostatically allowed to inject free carriers again, leading to a strong increase of the current. For V>VMAX1 it is even allowed to further fill-up the deep traps and also Qd will further increase. In that case the up-scan will start to connect to the up-

scan of the previous sweep. This process will continue for every subsequent sweep with a higher VMAX. From the observed values of VT1, VT2 and so on, we can

estimate the amount of deeply trapped electrons that stay behind in the device after each sweep. The total amount of charge carriers per area then increases from

Figure 6-2. Filling of shallow and deep traps during the up-scan of a J-V

measurement. During the back-scan the deep traps are not emptied, leading to

clockwise hysteresis.

s

d

Qf=0

Qs

Qd

- - - - -

Back scan

Up scan

V=0

Qf

Qs

Qd

V<VT1

J=0

-

-

- - - - -

VT1<V<VMAX1

--

-

- -

--

--

-

V=VMAX1

-Qf

Qs

Qd

s

d

Qf=0

Qs

Qd

- - - - -

Back scan

Up scan

V=0

Qf

Qs

Qd

V<VT1

J=0

-

-

- - - - -

VT1<V<VMAX1

--

-

- -

--

--

-

--

-

- -

--

--

-

V=VMAX1

-Qf

Qs

Qd

Chapter 6

78

1.4×1011 cm-2 after the first sweep to 4.2×1011 cm-2 after the 5th sweep. The average density of deeply trapped electrons per volume ntd is shown in Figure 6-3, and increases from 1×1016 cm-3 to 3×1016 cm-3 after 5 sweeps. It should be noted that

this concentration is at least an order of magnitude lower than the values of 5×1017 cm-3 estimated for the (effective) total amount of electron traps that follow from the

analysis of the (up-scan) J-V characteristics [2].

An important question is now whether these deeply trapped electrons are located in the bulk of the polymer, as sketched in Figure 6-1, or at the hole

blocking bottom electrode. Figure 6-4 shows the J-V characteristics of OC1C10-PPV based electron-only diodes made with Ga, In, and Yb bottom electrodes at

room temperature. The fact that the measured currents are nearly independent of the choice of the bottom electrode confirms the absence of chemical interactions between the bottom electrodes and the polymer. For voltages larger than 10 V the

devices with In and Ga show a slight enhancement of the current, that is accompanied by the onset of light-emission (not shown). This demonstrates that for

these electrodes hole injection starts to occur, and the measured currents are not solely due to the electrons anymore. The most important observation is that there is no change in the hysteresis behavior. This indicates that trapped electrons at the

bottom contact/polymer interface are not responsible for the observed hysteresis effects.

Figure 6-3. Concentration of charges remaining in deep traps after subsequent J-

V sweeps with increasing maximum applied voltage VMAX.

1 2 3 4 55.0x1015

1.0x1016

1.5x1016

2.0x1016

2.5x1016

3.0x1016

n td (cm

-3)

Nr of scan


79

In a recent study we demonstrated that in MEH-PPV the trap-limited currents can be strongly increased by addition of the n-type dopant

decamethylcobaltocene (DMC) [3]. By filling the traps with electrons from the DMC donor a trap- and hysteresis free space-charge limited electron current can be

obtained in MEH-PPV. As a next step we use a n-type doped PPV-based polymer as an interlayer between the metallic bottom electrode and the undoped MEH-PPV layer polymer for the fabrication of double-layer electron-only devices. In this case

the electron extracting electrode is not metallic, but a n-type doped polymer. However, in order to use such a n-type doped polymer layer as bottom electrode

the layer should not dissolve when the undoped MEH-PPV layer is spincoated on top of it. A way to circumvent the solubility issue is to tune the solubility by chemical modification [4]; PPV-based copolymers with selective solubility can be

achieved without loss of the charged transport properties. It was shown that by shortening the (2’-ethylhexyloxy) side chains, from poly[2,5-bis(2’ethylhexyloxy)-

1,4-phenylenevinylene] (BEH-PPV), to butoxy side chains the polymer poly[2,5-bis(butoxy)-1,4-phenylenevinylene] (BB-PPV) was obtained, which is only soluble

in chloroform in very low concentrations. Consequently, by tuning the ratio of the BEH- and BB- monomers the solubility could be adjusted over the whole spectrum of solvents. For example, the BEH-BB-PPV copolymer in a 1:3 ratio was only

soluble in chloroform, making it compatible with a large number of light-emitting polymers as MEH-PPV. In Figure 6-5 the J-V characteristics of an undoped BEH-

BB-PPV (1:3) and a DMC doped electron-only device is shown. The chemical

Figure 6-4. J-V characteristics of electron-only diodes with different hole blocking bottom electrodes (Metal/MDMO-PPV/Ba/Al).

0 10 20 30 4010-510-410-310-210-1100101102103

In Ga

J

(A/m

2 )

V (V)

Yb

Chapter 6

80

structure of BEH-BB-PPV (1:3) is shown in the inset of Figure 6-5. Similar as to earlier results on MEH-PPV the electron current increases a few orders of magnitiude, accompanied with an (almost) disappearance of the hysteresis.

This already demonstrates that the electrons from the dopants not

only fill the trap states that are responsible for the low electron currents, but also the trap states that are responsible for the hysteresis.

Figure 6-5. J-V characteristics of both an undoped BEH-BB-PPV (1:3) electron-

only (circles) and doped (squares) electron-only device.

0 5 10 15 20 2510-6

10-5

10-4

10-3

10-2

10-1

100

J (A/m

2 )

V (V)

L=365 nm

Figure.6-6. J-V characteristics of an electron-only device measured for a double

layer: doped BEH-BB1:3 and undoped MEH-PPV

0 2 4 6 8 1010-7

10-6

10-5

10-4

10-3

10-2

10-1

O

O*

O

O

*

13

J (A/m

2 )

V (V)

DopedL=220 nm

Undoped L= 230 nm

BEH/BB1/3


81

The electron current of the double-layer device, with an undoped MEH-PPV layer on top of the n-doped BEH-BB PPV layer is shown in Figure 6-6. It appears that addition of the MEH-PPV layer directly results in a large hysteresis. This clearly

shows that the hysteresis originates from electron trapping in the bulk of the undoped MEH-PPV.

In order to reduce or eliminate the hysteresis from the electron currents the deep traps have to be removed from the polymer material itself. In a recent study it

has been demonstrated that the optoelectronic properties of poly(3-hexylthiophene) (P3HT) can be strongly modified by treating the polymer solution with either

electrophiles or nucleophiles [5]. The p-type defects could be strongly reduced by a treatment with lithium aluminium hydride, whereas treatment with dimethylsulfate gives rise to a removal of anionic sites. To purify our crude MEH-PPV polymer, a

number of subsequent precipitations have been carried out. The precipitates where always filtered on a Whatman vacuumfiltration unit, using hydrophobic PTFE

membrane filters with a pore size of 0.45 µm from PALL Life Sciences. The crude MEH-PPV was dissolved in CHCl3 (5mg/ml) at 60 °C overnight under nitrogen

atmosphere. The solution was filtered to isolate the soluble polymer filtrate from the insoluble gel parts which stayed on the filter. A sample of the filtrate was subjected to analytical size exclusion chromatography (SEC), which was

performed using a Spectra series P100 (Spectra Physics) pump equipped with two mixed-B columns (10 µm, 2 cm x 30 cm, Polymer Labs) and a refractive index

detector (Shodex) at 70°C. THF was used as the eluent at a flow rate of 1.0 mL/min. Molecular weight distributions are given relative to polystyrene standards. The results can be found in table 1.

SEC Sample Mw PD

Unpurified MEH-PPV 236000 11.9

Reversed precipitation: Precipitate (High Mw MEH-PPV) 296000 3.2 Reversed precipitation: Filtrate

(Low Mw MEH-PPV) 19000 2.7

Table.1: Purification results for MEH-PPV before and after reversed precipitation,

determined by means of SEC in THF using polystyrene standards.

To decrease the polydispersity (PD) of the polymer mixture, cold methanol was added dropwise to the filtrate until the polymer starts to precipitate. Via this reverse precipitation, the larger polymer chains precipitate first while the shorter

Chapter 6

82

ones stay in solution. This way, only the high molecular weight polymer can be isolated by filtering the mixture. As can be seen in table 1, PD has decreased from 11.9 till 3.2 in the polymer precipitate. The SEC chromatograms of the 3 samples

are given in Figure 6-7. The high Mw polymer precipitate was purified further by two subsequent precipitations in respectively 500 ml acetone and 500 ml methanol.

The final polymer precipitate was collected and dried in vacuum overnight. The polymer is stored in dark under a nitrogen atmosphere.

After this purification procedure electron-only devices were again

fabricated. In Figure 6-8 the up- and down scan of the electron current for a

purified MEH-PPV is shown for a thickness of 300nm. We observe that after purification the hysteresis has completely disappeared. As a comparison also the J-

V characteristics of unpurified MEH-PPV is shown, with a film thickness of 340 nm. Here the electron current is lower and shows a clear hysteresis. The J-V characteristics of the purified MEH-PPV show the usual steep voltage dependence

and are modeled with a trap distribution that is exponentially distributed in energy. The parameters obtained are identical as the ones reported before [2]. An important

conclusion that can be drawn now is that the traps responsible for the trap-limited current in the purified MEH-PPV are not the same as the (deep) traps that are

responsible for the hysteresis. The exponential trap distribution is therefore represented by Qs in Figure 6-2. By now adding a single deep trap level with a concentration of 2×1016 cm-3 with a trap depth of ~0.7 eV we can describe the J-V

Figure 6-7. Overlay of SEC chromatograms of unpurified MEH-PPV (solid line),

high molecular weight MEH-PPV precipitate (dashed line) and low molecular

weight MEH-PPV filtrate (dotted line).

1000 10000 100000 1000000

0,0

0,2

0,4

0,6

0,8

1,0

Res

pons

e (a

.u.)

Molecular weight (Dalton)


83

characteristics of unpurified MEH-PPV reasonable well, in spite of the simple assumption of only a single trap level. The trap concentration agrees well with the estimation of the amount of trapped charges that stay behind in the device after a J-

V sweep, as shown in Figure 6-3. The deep traps responsible for the hysteresis correspond to Qd in Figure 6-2. The observation of two types of electron traps

corresponds with the earlier observation from thermally stimulated currents: in this study two main charge traps were found after excitation with light, with activation energies of 0.2-0.35 eV and 0.75-0.91 eV, respectively [6]. A possible origin for

the deep traps are hydrated oxygen clusters O2(H2O)n , which are located at around 3.7 eV below vacuum, leading to a 0.8 eV deep trap for MEH-PPV and MDMO-

PPV [7]. However, keeping our samples for 24 hours in vacuum (10-7 mbar) before deposition of the top-electrode did not lead to any improvement in the hysteresis behavior. Furthermore, the addition of Na2S04 to the polymer solution also did not

have any influence on the appearance of the hysteresis, suggesting that oxygen related defects might not be the origin.

Another possible origin for the deep traps can be assigned to the presence

of carbonyl containing end-groups in the polymer structure. Indeed detailed 13C-NMR studies have shown that Gilch PPV’s contain mainly aldehyde and

carboxylic acid end-groups [8]. These electron accepting functionalities could be reduced electrochemically via reaction with the injected electrons and thus act as

Figure 6-8. J-V characteristics of an electron-only device of the purified MEH-

PPV with L=300nm (squares) and an electron-only device not purified (circles)

together with the calculated currents for an exponential distribution (solid line)

and the addition of deep traps (dashed line).

0 5 10 15

10-6

10-5

10-4

10-3

10-2

J (A/m

2 )

V (V)

Chapter 6

84

deep traps). Purification leads to fractionation which allows to remove the lower molecular weight polymers and oligomers from the original batch. The resulting high molecular weight fraction will contain a strongly reduced number of said end-

groups possibly leading to a decrease of the number of deep traps in the polymer. The effect of the amount of low molecular weight fraction is a subject of further

investigations.

6.3 Conclusions

In conclusion, we demonstrate that the strong hysteresis observed in the

electron currents of PPV-based conjugated polymers can be attributed to the presence of deep traps. These traps are persistently occupied after a J-V sweep,

leading to clockwise hysteresis. The deep traps are not located at the interface between the polymer and the hole-blocking electrode, but are present in the

polymer itself. By proper purification of the PPV-based polymers hysteresis free electron currents can be obtained.


85

REFERENCES

[1] M. A. Lampert and P. Mark, Current Injection in Solids, Academic,

NewYork, 1970. [2] M. M. Mandoc, B. de Boer, G. Paasch, and P. W. M. Blom, Phys. Rev. B

75, 193202 (2007). [3] Y. Zhang, B. de Boer, and P. W. M. Blom, Phys. Rev. B. 81, 085201

(2010). [4] C. Tanase, J. Wildeman and P.W.M. Blom, Adv. Funct. Mater. 15 (2005)

2011.

[5] Z. Lang, A. Nardes, D. Wang, J. J. Berry, and B. A. Gregg, Chem. Mater. 21, 4914 (2009).

[6] V. Kazukauskas, H. Tzeng, and S.A.Chen, Appl. Phys. Lett. 80, 2017 (2002).

[7] J-M. Zhuo, L-H. Zhao, R-Q Png, L-Y Wong, P-J Chia, J-C Tang, S.

Sivaramakrishnan, M. Zhou, E.C.W. Ou, S-J. Chua, W-S Sim, L-L Chua, and P.K.H. Ho, Adv. Mater. 21, (2009).

[8] a. Becker, H.; Spreitzer, H.; Ibrom, K.; Kreuder, W. Macromolecules, 32, 4925-4932 (1999), b. H. Roex, P. Adriaensens, D. Vanderzande, and J. Gelan, Macromolecules 36, 5613 (2003).

Chapter 6

86

87

Publications

[1] N.I. Craciun, J. Brondijk and P.W.M. Blom, “Diffusion-enhanced hole

transport in thin polymer light-emitting diodes”, Phys. Rev. B. 77, 035206

(2008).

[2] N. I. Craciun, J. Wildeman and P. W. M. Blom, ‘Universal Arrhenius

temperature activated charge transport in diodes from disordered organic

semiconductors’, Phys. Rev. Lett. 100, 056601 (2008).

[3] N. I. Craciun1, J. Wildeman1, and P.W. M. Blom, ‘Substituted

polyfluorene based hole transport layer with tunable solubility’, J. Phys.

Chem. C 114, 10559 (2010).

[4] N. I. Craciun, Y. Zhang, A. Palmaerts, H. T. Nicolai, M. Kuik, R. J. P.

Kist, G. A. H. Wetzelaer, J. Wildeman, J. Vandenbergh, L. Lutsen, D.

Vanderzande, P. W. M. Blom, ‘Hysteresis-free electron currents in poly(p-

phenylene vinylene) derivatives’, J. Appl. Phys. 107, 124504, (2010).

Summary

89

Summary

Since the discovery of conducting polymers in 1977 the field of organic

electronics has evolved rapidly. Displays based on organic light emitting diodes

have already reached the consumer market and are produced in millions per month for applications as mobile phones. The development of organic solar cells and integrated circuits is still steadily increasing. One of the great benefits of these

‘plastic’ electronic materials is that they can be processed from solution. This enables cheap and fast production of electronic devices using roll-to-roll based

technology, similar as the production of newspapers. Deposition technologies as slot-die coating and ink jet printing enable the production organic electronic devices on unconventional substrates, such as flexible plastic foils and paper.

Another great benefit is the possibility of tailoring the polymers through carefully controlled synthesis, resulting in a multitude of different functionalities. Since the

discovery of electroluminescence in conjugated polymers in 1990 it has been recognized that charge transport is a key ingredient for the efficiency of the

polymer light-emitting diodes (PLEDs). The active part of PLEDs usually consists of only a single layer. In such a device holes are injected from the anode and electrons from the cathode into the polymer layer. The use of only a single electro-

optic layer has a large fundamental disadvantage: due to the reduced electron transport in conjugated polymers most of the light in a PLED is generated close to

the metallic cathode. This metallic cathode acts as a quenching site for the generated excitons, thereby strongly reducing the efficiency of the PLEDs. In order to strongly increase the efficiency of PLEDs multilayer stuctures have to be

developed, comparable to the heterojunction based LEDs and laser diodes in inorganic III-V semiconductors. In such an optimized device electrons and holes

are efficiently transported via high mobility layers towards a highly luminescent layer. The energy levels of the polymers used will be designed in such a way that the hole transport layer also serves as a blocking layer for electron transport and

vice versa. Using a heterostructure electrons and holes can not leave the device without recombining, leading to an enhanced efficiency. A major problem why

polymer based multilayer devices have not been realized so far is the solubility of the materials used; a multilayer can not be fabricated when a spin casted layer dissolves in the solvent of the subsequent layer. In this thesis charge transport

layers with compatible solubility are developed and electrically characterized. The transport of holes in polymer light-emitting diodes (PLEDs) based on poly (2-

methoxy, 5- (2’ ethyl-hexyloxy)-p-phenylene vinylene) (MEH-PPV) is investigated as a function of layer thickness. For thicknesses smaller than 100 nm

Summary

90

the current in these thin PLEDs is strongly enhanced as compared to the expected space-charge limited (SCL) current. Applying the standard SCL model to

measurements on a PLED with a thickness of only 40 nm results in an apparent increase of the hole mobility of a factor of 40. We show that this strong increase of

the hole transport properties in these thin devices originates from the presence of an Ohmic hole contact. For Fermi-level alignment holes diffuse from the contact into the MEH-PPV, forming an accumulation layer with a width of a few tens of

nanometers. Due to the density dependence of the mobility the hole transport in this accumulation region is strongly enhanced. For the analysis of thin PLEDs it is

therefore essential that both drift and diffusion of charge carriers are taken into account. Charge transport models developed for disordered organic

semiconductors predict a non-Arrhenius temperature dependence ln(µ) ∝1/T2 for

the mobility µ. We demonstrate that in space-charge limited diodes the hole mobility (µh) of a large variety of organic semiconductors shows an universal

Arrhenius temperature dependence µh(T)=µ0exp(-∆/kT) at low fields, due to the presence of extrinsic carriers from the Ohmic contact. The transport in a range of organic semiconductors, with a variation in room temperature mobility of more

than six orders of magnitude, is characterized by an universal mobility µ0 of 30-40 cm2/Vs. As a result we can predict the full temperature dependence of their charge

transport properties with only the mobility at one temperature known. We report on the synthesis and electrical characterization of polyfluorene-

triarylamine based hole transport layers (HTL). The solubility of the HTL can be

tuned by adjustment of the chemical structure, without loss of the charge transport properties. Double-layer polymer light-emitting diodes are constructed with a HTL

that is not soluble in toluene at room temperature and as light-emitting layer MEH-PPV is used. The addition of the HTL enhances the efficiency of the PLED with 10% at higher voltages.

The interpretation of electron currents in conjugated polymers is strongly hindered by the occurrence of hysteresis. We investigate the transport of electrons in

electron-only devices based on derivatives of poly(p-phenylene vinylene) (PPV) for various hole-blocking bottom electrodes as well as purification of the polymer.

The use of a variety of hole blocking bottom contacts, as metallic electrodes and n-type doped polymers, did not give any improvement in the observed hysteresis. By purification of the PPV, hysteresis free electron-only currents can be obtained. The

deep traps responsible for hysteresis, with a concentration in the 1016 cm-3 range, are not responsible for the trap-limited electron transport as observed in purified

PPV.

91

Samenvatting

Na de ontdekking van geleidende plastics in 1977 zijn de ontwikkelingen

in de organische elektronica heel snel gegaan. Beeldschermen gebaseerd op organische lichtemitterende diodes zijn al op de markt en worden in grote aantallen

van enkele miljoenen per maand geproduceerd voor toepassing in o.a. mobiele telefoons. Ook de ontwikkeling van organische zonnecellen en geïntegreerde

circuits vordert gestaag. Een groot voordeel van de elektronische plastics is dat ze vanuit vloeistof op oppervlakken aangebracht kunnen worden. Hierdoor wordt het mogelijk om elektronische devices goedkoop en op hoge snelheid te produceren

door middel van productie aan de rol, identiek aan het maken van kranten. Het deponeren van actieve elektronische lagen via slot-die coaten en ook ink-jet printen

maakt het ook mogelijk om elektronische devices op andere dan silicium substraten te maken, zoals plastic folies en papier. Verder kan de chemische structuur van polymeren door middel van de synthese zodanig aangepast worden

dat het polymeer allerlei nieuwe functies kan uitvoeren. Na de ontdekking van elektroluminescentie in geconjugeerde polymeren in 1990 was het al duidelijk dat

het transport van ladingen in het polymeer een belangrijk proces is dat mede de efficiëntie van een polymeer lichtemitterende diode (PLED) bepaalt. Het actieve gedeelte in een PLED bestaat gewoonlijk uit een enkele lag. Vanuit de kathode

worden dan elektronen in deze laag geïnjecteerd en vanuit de anode gaten. Het feit dat er maar een enkele actieve laag is heeft een groot nadeel: omdat het elektronen

transport in de PLED veel slechter is dan het gaten transport wordt het meeste licht vlak bij de kathode gegenereerd. Aangezien de elektron-gat paren die in de PLED

gegenereerd worden hun energie aan de metallische kathode kunnen overdragen in plaats van licht uit te zenden leidt dit tot verliezen. Deze verliezen kunnen beperkt worden door niet een enkele maar meerdere organische lagen te gebruiken in de

PLED, net zoals ook in anorganische LEDs, gebaseerd op III-V halfgeleiders, toegepast wordt. In zo een LED worden de gaten en elektronen via lagen met

goede transporteigenschappen aangevoerd naar een emitterende laag met goede optische eigenschappen. De energie niveaus van de verschillende materialen worden dan zo gekozen dat de gaten transport laag ook de elektronen tegenhoudt

en omgekeerd. Door gebruik te maken van een heterostructuur kunnen de gaten en elektronen niet door de LED getransporteerd worden zonder daarbij te

recombineren, hetgeen tot hoge efficiënte devices leidt. Een groot probleem bij het maken van meerlaags structuren van geleidende polymeren is het feit dat ze oplosbaar zijn. Als de ene laag oplost als de volgende er op wordt aangebracht is

het niet mogelijk een meerlaags LED te maken. In dit proefschrift wordt de

Samenvatting

92

ontwikkeling en elektrische karakterisatie van polymeren laagjes met een geschikte oplosbaarheid die gaten en elektronen kunnen transporteren beschreven. Het transport van gaten in heel dunne laagjes (40-300 nm) van het

geconjugeerde polymeer poly (2-methoxy, 5- (2’ ethyl-hexyloxy)-p-phenylene vinylene) (MEH-PPV) is onderzocht. Het blijkt dat de transporteigenschappen in

hele dunne lagen (<100 nm) beter zijn dan zoals verwacht uit de metingen aan dikke lagen en zelfs de ruimteladings-begrensde limiet overschrijden. Voor lagen van 40 nm dik lijkt de gatenmobiliteit zlefs met een factor 40 toe te nemen. We

laten zien dat dit wordt veroorzaakt door de nabijheid van een Ohms contact. Vanwege de gradiënt in de ladingsdragersdichtheid tussen de elektrode en het

polymeer diffundeert er in het polymeer, typisch in een gebied van enkele tientallen nanometers. Omdat het transport in een polymeer afhangt van ladingsdragersdichtheid wordt in dit gebied het transport van ladingen sterk

verbeterd. Het is daarom belangrijk dat bij de analyse van het ladingstransport in dunne lagen dit diffusie effect meegenomen wordt. De modellen voor

ladingstransport in geleidende polymeren en andere wanordelijke materialen

voorspellen een temperatuurafhankelijkheid van de mobiliteit volgens ln(µ) ∝1/T2. Echter, wij laten zien de mobiliteit van gaten in een groot aantal organische

geleiders een universele Arrhenius achtige temperatuursafhankelijkheid laat zien, gegeven door µh(T)=µ0exp(-∆/kT). Dat deze schaling optreedt kan veroorzaakt

worden door de extra ladingsdragers die uit het contact in de polymere halfgeleider zijn gediffundeerd. Het transport van een serie van organische halfgeleiders, waarvan de mobiliteit bij kamertemperatuur zes orders van grootte verschilt, wordt

gekarakteriseerd door een universele mobiliteit µ0 van 30-40 cm2/Vs.

Dientengevolge kunnen we van ieder polymeer de temperatuursafhankelijkheid

van het transport voorspellen door alleen maar de mobiliteit bij kamertemperatuur te meten. Vervolgens rapporteren we over de synthese en elektrische karakterisatie

van een nieuw polymeer, gebaseerd op polyfloureen triarylamine, welke als gatentransport laag (GTL) in meerlaags PLEDs gebruikt kan worden. De

oplosbaarheid van de GTL kan aangepast en ingesteld worden door middel van de chemische structuur zonder dat de ladingstransporteigenschappen veranderen. De

PLEDs bestaan dan uit een GTL die niet in tolueen oplost met daar bovenop MEH-PPV als lichtemitterende laag. Deze twee-laags PLED heeft bij hogere spanningen een efficiëntie die 10% hoger is dan een enkellaags device.

Het meten van het elektronentransport in geconjugeerde polymeren wordt sterk bemoeilijkt door het optreden van hysterese tijdens de metingen. We hebben

het elektronen transport in MEH-PPV onderzocht door gebruik te maken van verschillende onderelektrodes als ook verschillende zuiveringen van het polymeer. De variaties in de onderelektrode, va metalen tot n-type gedoteerde polymeren,

Samenvatting

93

hadden geen enkel effect op het optreden van hysterese. Echter, door te zuiveren waren we in staat de hysterese vrijwel volledig te elimineren. De diepe traps die de hysterse veroorzaken, typisch met een concentratie van 1016 cm-3 , zijn echter niet

verantwoordelijk voor het gereduceerde elektronen transport zoals dat in het gezuiverde polymeer optreedt.

Samenvatting

94

95

Acknowledgements

The research presented in this thesis has been carried out at the Molecular

Electronics- Physics of Organic Semiconductors Laboratory from the University of

Groningen, The Netherlands, and was financially supported by the Zernike

Institute for Advanced Materials and the Dutch Technology Foundation (STW). I

would like to gratefully acknowledge several people whose contribution was

essential for the accomplishment of this thesis.

First, I thank my promoter, Paul Blom, for giving me the opportunity to pursue

a Ph.D. in his group. Throughout all these years, you were always a friendly leader,

you gave me the best advice, explanation and supervision whenever necessary and

showed lot of patience with me. Hartelijk bedankt!

I would also want to thank the members of the reading committee, Dago de

Leeuw, Reinder Coehoorn and Dirk Vanderzande for taking the time to read the

thesis and for their remarks and suggestions.

The work described in this thesis would not have been possible without all the

materials from Jurjen Wildeman. Jur, you always had “a good feeling about this”

and you always succeeded to bring me in a good mood. Thank you for all the time

you took in making new materials, and for our fruitful discussions and

collaboration.

I would also want to thank our secretaries for all the administrative help and

special thanks to Renate for all her help regarding the papers for this thesis.

The experiments presented in this thesis could not have been realized without

the excellent technical support in respect to all kinds of issues.

The result of my work would not have been possible without help and support

from the group members: André for teaching me how to use the measurement set-

up, Johan for all the nice samples we made together, Magda for teaching me how

to make my first LED and for all the nice discussion we had during the years, Jan

Anton for ‘SIMsalabim’, Herman for teaching me how to make ‘electron-only’s

and all the help with data modeling, Martijn, René and Milo for the nice

collaboration, Afshin, Hylke and Jia for the nice ‘smoking breaks’, Ilias for all the

nice CD’s I got as a ‘souvenir’ and for letting me know about every nice concert

(thank you!!!), and of course the rest or the MEPOS group for all the nice time we

spent together.

Samenvatting

96

Special thanks goes to Dan Crînguş for unlocking my door when I got stuck or

technical support with the thesis. Mulţumesc frumos!

Last I would like to thank Francesco for listening and being there whenever I

needed someone to talk and Gert-Jan for all his help and support in the last year

and for making me the cover that I wanted, but much prettier.

University of Groningen Electrical characterization of polymeric charge transport ... ·...

Documents

Transcript of University of Groningen Electrical characterization of polymeric charge transport ... ·...