Efficiency Roll-Off in Organic Light-Emitting Diodes

TL;DR: This review summarizes the current knowledge and provides a detailed description of the relevant principles for exciton-quenching mechanisms, both for phosphorescent and fluorescent emitter molecules, and further review methods that may reduce the roll-off and thus enable OLEDs to be used in high-brightness applications.

Abstract: Organic light-emitting diodes (OLEDs) have attracted much attention in research and industry thanks to their capability to emit light with high efficiency and to deliver high-quality white light that provides good color rendering. OLEDs feature homogeneous large area emission and can be produced on flexible substrates. In terms of efficiency, OLEDs can compete with highly efficient conventional light sources but their efficiency typically decreases at high brightness levels, an effect known as efficiency roll-off. In recent years, much effort has been undertaken to understand the underlying processes and to develop methods that improve the high-brightness performance of OLEDs. In this review, we summarize the current knowledge and provide a detailed description of the relevant principles, both for phosphorescent and fluorescent emitter molecules. In particular, we focus on exciton-quenching mechanisms, such as triplet-triplet annihilation, quenching by polarons, or field-induced quenching, but also discuss mechanisms such as changes in charge carrier balance. We further review methods that may reduce the roll-off and thus enable OLEDs to be used in high-brightness applications.

Summary

1 Introduction

• For more than 100 years, this inefficient but low-cost light source could not have been imagined to vanish from their daily life.
• OLEDs are furthermore attractive for lighting due to their potential to achieve good color rendering.
• The origin of efficiency roll-off is still under discussion. [11].
• The investigations are extended to research the influence of the host material and of the processing parameters.

2 Principles of Organic Semiconductors

• Organic semiconductors offer interesting properties that can be used in state-of-the-art electronic devices.
• One of the interesting properties of organic semiconductors is that the optical, electrical, and structural properties can be tailored to special purposes by chemical engineering. [14].
• In the example of the molecule ethene, the molecular sp2- and pz-orbitals of the carbon atoms split into bonding and anti-bonding σand π-states (see Fig. 2.3).
• In order to obtain Ohmic charge injection and increase the conductivity, electrical doping is used.
• This is especially important for phosphorescent emitter molecules in order to avoid exciton quenching due to high guest concentration. [39].

3 Theory of Efficiency Roll-Off

• Efficiency Roll-Off describes the efficiency loss of an OLED with increasing current density.
• In the following, the influence and relative importance of the processes introduced above is discussed.
• 4 Scope of this Work As discussed in the previous section, TTA is the dominating annihilation process underlying efficiency roll-off in most phosphorescent OLEDs.

4 Experimental Methods

• This chapter briefly describes the experimental methods.
• Materials are evaporated onto 1 mm thick glass substrates that are cleaned prior to use.
• The following procedure is applied in order to measure transient photoluminescence (see Fig. 4.3): Samples are excited with a nitrogen laser (MNL 200, Lasertechnik Berlin) at 337 nm.
• There- 36 experimental methods fore, a spectro-goniometer is used to measure the angular dependent spectra.
• The emitter compounds are doped into matrix materials in order to avoid concentration quenching. [39].

5 Influence of the Optical Environment

• The exciton density and, hence, the efficiency roll-off depend on the emitter lifetime, which can be influenced by the optical environment around the emitting dipoles.
• A clear correlation between roll-off and triplet lifetime is observed for both emitters, i.e., the data extracted from J0 are in good agreement with the fits obtained from transient electroluminescence measurements.
• A function of ETL thickness) are combined with calculations of the efficiency roll-off, using the Ir(MDQ)2(acac)-based OLEDs (Stack A) as an example.
• Therefore, the first maximum device is more efficient at high current densities than the second maximum device.
• Figure 5.8b shows the calculated critical current density J0 of the three devices in dependence of the ETL thickness dETL.

6 Influence of the Emission Profile

• The spatial exciton distribution inside the emission layer is described by the emission profile of an OLED.
• This chapter focusses on the derivation and modification of the emission profile in order to decrease the local exciton density and, thus, improve the efficiency roll-off.
• Deviations between 6.1. preliminary considerations (b) Critical current density J0 in dependence of the emission zone width w taking TTA (solid line) or TTA and TPA (dashed line) as loss channels into account.
• MixedEMLs (M-EML) consisting of a single layer that comprises a mixture of a hole and an electron transporting matrix material provide a broader emission zone and, thus, improve the roll-off further.
• The sensing profile may now be derived from S(x0) = I − ISL(x0) I , (6.6) where I is the intensity of the reference sample, and ISL(x0) is the signal from the excitons in the sensing layer device that are not quenched.

M-EML HBLEBL

• EBL/EML/HBL structure of M-EML samples using doped blockers, also known as Figure 6.25.
• The negative intensity observed in the green spectral part for samples with doped EBL could be related to contamination of the reference samples with the green emitter during evaporation.
• The measurement of the emission profile and the investigations using doped blockers indicate that the emission profile in M-EML devices might indeed be broader compared to D-EML, which is caused by the less confined charge exciton formation region.
• Finally, the measured sensing intensities are fitted allowing extraction of the exciton generation profile, the diffusion length, and the shape and width of the emission profile.
• Therefore, independent measurement of the charge balance and, especially, the avoidance of charge imbalance should be studied in more detail.

7 Influence of Molecular Aggregation

• The triplet-triplet annihilation rate constant, which is one of the factors determining the strength of efficiency roll-off, can be altered by aggregation of emitter molecules.
• Throughout this chapter, the influence of the molecular properties of the emitter is studied in more detail.
• For details of the measurement 7.2. aggregation of homoleptic and heteroleptic emitters Hatched squares indicate the TTA regime.
• Figures 7.11a and b show the results for Ir(ppy)3 and Ir(ppy)2(acac), respectively, as well as for their mixture into TCTA at different concentrations.
• In contrast to their report, molecular orientation is here especially observed for pure emitter layers and seems to be an intrinsic property of the material growth.

8 Summary and Outlook

• This chapter summarizes the results of the thesis and describes the interplay of optical environment, emission profile, and emitter aggregation with regard to efficiency roll-off.
• This enables an enhancement of efficiency at high brightness for a broad range of emitters and OLED structures.
• Efficiency roll-off is not only based on high exciton densities, but also on many other mechanisms.
• For both concepts, the triplet lifetime could be modified with the method proposed in Chapter 5.
• Further possibilities include evaporation onto tilted substrates and systematic variation of the surrounding host molecules.

Figures

Figure 6.3: Illustration of the emission zone width w compared to an exponential emission profile. The overall exciton density is kept constant.

Table 7.5: Selected properties of phosphorescent emitters. Values of the emitter’s diameter R and the dipole moment µD are taken from Ref. 99. The dipole-dipole potentials between two emitter molecules UGG and between emitter and host UGH are calculated relative to the potential of Ir(ppy)3 using Eq. 7.9. Homoleptic emitters are marked with a blue background.

Figure 2.5: Illustration of the Kasha rule. Absorption (red) and emission (blue) possess a mirror symmetry, shifted by the Stokes shift (green arrow).

Figure 3.3: Illustration of spinstatistics in electrically driven OLEDs. Figure adapted from Ref. 108.

Figure 3.2: Schematic illustration of possible mechanisms leading to efficiency roll-off in OLEDs. The different particles involved are electrons (e−), holes (h+), singlets (S, red), and triplets (T, green). Connecting lines indicate whether a particle is destroyed (open circle), created (full circle), or preserved (no circle). The mechanisms involved are charge imbalance, bimolecular quenching processes such as TPA/SPA, SSA, STA, TTA, exciton dissociation under the influence of heat or an electric field, and the outcoupling of the created photons. Mechanisms that might depend on current density and would thus influence the rolloff are shown in italics. For completeness, exciton formation (EF), intersystem crossing (ISC), and photon formation (PF) are also shown.

Figure 4.5: Chemical structure and PL spectrum of the blue emitter FIr6.

Figure 7.13: (a) Powder pattern of Ir(ppy)3 with indication of the four main peaks. (b) Packing diagram of Ir(ppy)3 showing the unit cell containing eight molecules. Molecules are outlined by their iridium cores and the three nitrogen atoms. The spatial depth of the molecules is indicated by decreasing color intensity. (c) Four unit cells with tetramers of Ir(ppy)3, displayed along c. Large/small circles correspond to higher/lower lying Ir(ppy)3 molecules. (+/−)- signs indicate the direction of the permanent dipole moment of the molecule, which is approximately parallel to the c-axis. Crystallographic data for (a,b) and the figure of part (c) are taken from Ref. 264. Reprinted with permission. Copyright 2010, Wiley VCH.

Figure 2.1: Molecular orbitals of ethene. Overlapping pz orbitals form the π-bond. H-atoms (omitted for clarity) lie in the plane and are connected to the C-atoms with σ-bonds.

Figure 5.5: Transient electroluminescent intensity data of two typical samples based on Ir(MDQ)2(acac) after excitation with a 50 µs long voltage pulse of 2.5 V (points). The data are fitted to a monoexponential decay function (including a constant background) to extract the triplet lifetime τ⋆ (lines).

Figure 2.11: Energy levels and transitions in a phosphorescent hostguest system upon electrical excitation.

Table 7.3: Doping concentration Γ of mixed TCTA:guest layers in mol % and fit results from XRR measurements yielding the layer thickness d and roughness ξ.

Figure 7.19: Influence of substrate temperature on emission of TCTA:Ir(ppy)3 at 9 mol %. (a) Normalized PL intensity and (b) triplet exciton density nT as a function of the pump exciton density governed by time-resolved PL measurements. and is therefore omitted. [287] All samples are fabricated on one substrate within one fabrication run. Evaporation is started at the highest temperature in order to avoid heating of already prepared samples. Figure 7.19a shows the normalized PL spectra of all samples. A very small blue-shift with increasing temperature is observed, but the overall effect is marginal. The absolute intensity instead stays constant up to 75 ◦C and then decreases with higher temperatures (see Fig. D.2a in App. D). This could in fact hint to aggregation and exciton annihilation, respectively.

Figure 6.20: Structure of the investigated OLEDs: (a) D-EML and (b) M-EML device. For M-EML, the ratio of the two matrix materials x : y is varied.

Figure 6.24: Measured external quantum efficiency (dots) and fits according to Eq. 3.4 (lines) for D-EML (solid line) and M-EML (dashed lines) with varying matrix ratio.

Table 6.2: Critical current density J0 extracted from fits of Eq. 3.4 to the EQE data shown in Fig. 6.24.

Table 7.6: Doping concentrations Γ1 and Γ2, and the relative dipoledipole potential UGH/UIr(ppy)3 between guest and host of all investigated host-guest systems. Blue backgrounds mark the systems that have been studied in Section 7.2.

Figure 7.16: Spectroscopic parameters of (a,d) Ir(ppy)3, (b,e) Ir(ppy)2(acac), and (c,f) Ir(MDQ)2(acac) doped into different host materials. Data are accessed by fitting the emission intensity I(E)/E3 to the Poisson progression in Eq. 7.3. (a-c) Absolute values at Γ1 doping concentration (see Table 7.6). The parameters E00, ⟨E⟩, FWHM, h̄ω, and σν are given in eV; s is dimensionless. (d-f) Relative change ∆X of the parameters with increasing doping concentration. Values are calculated according to Eq. 7.5 and are given in % per mol %. Blue bars mark the hosts that were used in Sec. 7.2.

Figure 2.10: Energy levels of a typical OLED structure applying forward bias between anode and cathode.

Figure 2.9: Schematic illustration of band bending at the interface between metal and organic when introducing a p-dopant.

Figure 6.5: Normalized EQE vs. current density for emission zones of different width w under the TTA model (solid lines) or the combined TTA/TPA model (dashed lines). The roll-off is shifted to higher current densities for broader emission zones. The blue dotted curve represents an assumed broadening of the emission profile with increasing current density from 2 to 20 nm. For this case, the EQE vs. current density curve shows an s-kink at a current density of approximately 10 mA/cm2 (red lines).

Figure 2.2: Schematic illustration of two electronic transitions E(R) and their vibrational structure ν.

Full text

C A R O L I N E M U R A W S K I
Geboren am 22.02.1987 in Dresden,
Geburtsname: Caroline Weichsel
Efficiency Roll-Off in
Organic Light-Emitting Diodes
Dissertation zur Erlangung des akademischen Grades
D O C T O R R E R U M N AT U R A L I U M
T E C H N I S C H E U N I V E R S I TÄT D R E S D E N

Technische Universität Dresden
Fakultät Mathematik und Naturwissenschaften
Fachrichtung Physik
Institut für Angewandte Photophysik
Eingereicht am 21.04.2015
Verteidigt am 28.08.2015
1. Gutachter: Prof. Dr. Karl Leo
2. Gutachter: Prof. Dr. Malte C. Gather

In Erinnerung an meinen Großvater Prof. Dr. Ludwig Walther

Abstract
The efficiency of organic light-emitting diodes (OLEDs) typically
decreases with increasing current density. This so-called roll-off im-
pedes the market entry of OLEDs in high-brightness applications
such as general lighting. One of the most important processes causing
roll-off is exciton annihilation, which evolves upon high exciton den-
sities. This mechanism is especially pronounced in phosphorescent
molecules due to their long triplet lifetime. In order to reduce the
roll-off in phosphorescent OLEDs, this thesis focusses on decreasing
the local exciton density by modifying the exciton lifetime, the spatial
exciton distribution, and the tendency of emitters to form aggregates.
The obtained results lead to a deeper understanding of efficiency
roll-off and help sustaining the OLED efficiency at high brightness.
The emitter lifetime can be influenced by the optical environment
around the emitting dipoles through the Purcell effect. In order to
study this effect, the distance between emitter and metal cathode is
varied for two different OLED stacks. A strong influence of emitter
position and orientation on roll-off is observed and explained by
modelling the data with triplet-triplet annihilation theory. Further-
more, design principles for optimal high-brightness performance are
established by simulating the roll-off as a function of emitter-cathode
distance, emissive dipole orientation, and radiative efficiency.
Next, a method is developed that allows extracting the spatial exci-
ton distribution. Therefore, a thin sensing layer that locally quenches
excitons is introduced into the emission layer at varying positions.
The resulting quenching profile is then fitted using a comprehensive
theory based on the diffusion equation, which renders the exciton
distribution and diffusion length with nanometer resolution. This
method is applied to an emission layer comprising an ambipolar host
material. Contrary to expectations which suggest that ambipolar
materials exhibit broad exciton formation, a narrow emission zone
close to the electron transport layer is found. Additional explorations
of structures that might broaden the emission zone point to a nar-
row emission zone in double emission layers and broader exciton
formation in mixed emission layers.
Previous investigations revealed a strong correlation between emit-
ter aggregation and molecular dipole moment of the emitter. Within
this thesis, the range of studied emitters is significantly extended.
It is shown that homoleptic emitters show a stronger tendency to
form aggregates than heteroleptic compounds. This is probably not

Frequently asked questions

Q1. What are the contributions mentioned in the paper "Efficiency roll-off in organic light-emitting diodes" ?

Within this thesis, the range of studied emitters is significantly extended. 

Q2. How are the design principles for the optimal high brightness performance developed?

By further simulating the roll-off as a function of emitter-cathode distance, emissive dipole orientation, and radiative efficiency, design principles for optimal high-brightness performance are developed. 

Q3. Why are electrons more strongly bound in the -bond?

Due to the larger overlap of the sp2-orbitals compared to the pz-orbitals, electrons are more strongly bound in the σ-bond than in the π-bond. 

Q4. What is the effect of TTA on the efficiency of phosphorescent OLEDs?

Depending on the excited state lifetime, current density, and, hence, on the triplet density, TTA may significantly decrease the efficiency of phosphorescent OLEDs. 

Q5. What is the effect of the ratio of electrons and holes in the EML?

if the ratio of electrons and holes in the EML is not balanced, for instance by charge carrier built-up at interfaces or bad injection for either electrons or holes, the electrical efficiency will decrease. 

Q6. How can the effect of the molecular structure on efficiency roll-off be studied?

7.2.2 Time-Resolved SpectroscopyUsing time-resolved spectroscopy, the effect of the molecular structure on efficiency roll-off can be studied by investigating the strength of triplet-triplet annihilation in the respective materials. 

Q7. How does the disorder in organic semiconductors affect the charge transport?

[50–53]Compared to inorganic semiconductors, where electronic bands are formed leading to charge carrier mobilities of approximately 103 cm2/(V s), the disorder present in organic semiconductors strongly12 principles of organic semiconductorshinders charge transport. 

Q8. What is the use of specular geometry in OLEDs?

The specular geometry is here only used for X-ray reflection (XRR) measurements, which allow evaluation of film thickness and roughness. 

Q9. How did Giebink and his colleagues study the importance of charge balance effects for a?

Using time-resolved experiments, the authors investigated the importance of charge balance effects for a range of OLED structures. 

Q10. How many times has a value of kTT been obtained for a similar OLED?

For the latter, a value of kTT = (3 ± 2)× 10−12 cm3/s has previously been obtained for a similar OLED stack using transient decay measurements at high excitation densities. [13] 

Q11. What could be used to determine the anisotropy factor of the triplet state?

measurement of the roll-off for different OLED microcavities and concurrent optical modelling could be used to determine the anisotropy factor. 

Q12. What is the effect of the decay rate of the emitter on the efficiency of the OLED?

This is related to the increase of ηrad in decay rate at small ETL thickness (cf. Fig. 8 in Ref. 68): If the radiative efficiency is high, the total decay rate of the emitter is dominated by the effective radiative decay rate, which in turn strongly increases in close proximity to the metal contact, thus reducing the roll-off.