Light-Emitting Electrochemical Cells
Rubeacuten D Costa Editor
Concepts Advances and Challenges
Light-Emitting Electrochemical Cells
Rubeacuten D CostaEditor
Light-EmittingElectrochemical CellsConcepts Advances and Challenges
123
EditorRubeacuten D CostaIMDEA MaterialesParque Cientiacutefico yTecnoloacutegico-Tecnogetafe
Getafe (Madrid)Spain
ISBN 978-3-319-58612-0 ISBN 978-3-319-58613-7 (eBook)DOI 101007978-3-319-58613-7
Library of Congress Control Number 2017940608
copy Springer International Publishing AG 2017This work is subject to copyright All rights are reserved by the Publisher whether the whole or partof the material is concerned specifically the rights of translation reprinting reuse of illustrationsrecitation broadcasting reproduction on microfilms or in any other physical way and transmissionor information storage and retrieval electronic adaptation computer software or by similar or dissimilarmethodology now known or hereafter developedThe use of general descriptive names registered names trademarks service marks etc in thispublication does not imply even in the absence of a specific statement that such names are exempt fromthe relevant protective laws and regulations and therefore free for general useThe publisher the authors and the editors are safe to assume that the advice and information in thisbook are believed to be true and accurate at the date of publication Neither the publisher nor theauthors or the editors give a warranty express or implied with respect to the material contained herein orfor any errors or omissions that may have been made The publisher remains neutral with regard tojurisdictional claims in published maps and institutional affiliations
Printed on acid-free paper
This Springer imprint is published by Springer NatureThe registered company is Springer International Publishing AGThe registered company address is Gewerbestrasse 11 6330 Cham Switzerland
Foreword
Light-Emitting Electrochemical Cells organic semiconductordevices augmented by ions
Polymer light-emitting electrochemical cells (LEC) was invented in 1994 tofacilitate the injections of charge carriers into luminescent conjugated polymers inlight emitting diodes Initially the benefits of adding a solid electrolyte into theorganic semiconductor devices was eminent electrons could be effectively injectedfrom a stable metal like aluminum into the lowest unoccupied molecular orbitalswhich are 1 eV higher than aluminumrsquos work function and the driving voltage wassubstantially reduced Also the emissive layer could be much thicker than thatallowable for tunneling charge injection interfaces Michael Rubner at MITobtained similar results with conjugated small molecules such as a soluble
derivative of RuethbpyTHORN3 2thorn PF6
2 a compound that had been studied for eletro-
generated chemiluminescence long before LEC was first reported The presence ofmobile ions in the LECs and consequently overpotential could induce degradationwhich has been partially addressed over the years Karl Leo and Junji Kido et alintroduced immobile ions or ldquodopedrdquo the organic semiconductor via organicdonor-acceptor complex which have leap frogged the operational stability ofOLEDs
Works by Richard Friend Ian Parker and others in the early 1990s showedthat the conjugated polymerelectrode interfaces could be modeled as tunnellingbarriers The work function difference between ITO a commonly used transparentanode and calcium is around 18 eV Low electron and hole injection barriers arefeasible when using a small band gap polymer such as MEH-PPV to produceorange light emission For blue light emitting polymers large barriers are inevi-table The emissive layer has to be made as thin as processing can produce adefect-free layer typically around 100 nm for spin coating Even so the drivingvoltage was often tens of volts and the blue polymer OLEDs fabricated at the time
v
were very short lived even though the quantum efficiency was decent In finding asolution to this challenge the doping propagation model that Olle Inganaumls and meused to simulate the polypyrrolepolyethylene bilayer bending beams inspired me tointroduce electrochemical doping into the polymer OLEDs Thus a commonly usedionically conductive polymer polyethylene oxide plus lithium triflate was selectedto supply the mobile dopants The Wessling precursor of PPV was selected thanksto its compatibility with the PEO-lithium salt The resulting blend of an ionicallyconductive and an electronically conductive polymers showed remarkablyimproved electroluminescent performance compared to control OLEDs based onITOPPVAl The driving voltage was lower the quantum efficiency was higherand the operational stability was also enhanced
Alan Heeger Jun Gao Ludvig Edman and others separately confirmed theformation of p-i-n junction in the polymer LECs by optical beam induced currentmeasurements direct imaging under microscope and measurement of electrostaticpotential distribution by scanning Kelvin probe microscopy Light emission andmajor potential drop were observed to occur at the junction This p-i-n junctionmodel may not rule out other mechanisms particularly when the junction is notformed to exhibit the ideal doping profiles at the electrode interfaces Electricaldouble layers could dominate at driving voltage well below the band gap of theconjugated polymer ie the onset of simultaneous p- and n-doping
The junction model essentially guides the material selection to fabricate highperformance LECs Three electronicionic polymer blend systems were examined inthe early years include (MEH-PPV + PEO-lithium salt) and (polyfluorene withethylene oxide side chains + lithium salt) in addition to the (PPV + PEO-LiTf)system used in the very first LEC device Morphological control was a criticalfactor in the device performance one had to consider the tradeoffs among carriertransport ionic mobility luminescence quenching and accessibility of doping ionsinto the low-polarity conjugated polymer domain Furthermore mobile ions couldlead to electrochemical over-reaction or degradation at high driving voltagesFreezing the ions after the formation of the p-i-n junction appears to be effective toslow down such degradation
The added freedom in electrode selection for LECs allows the fabrication ofdevices without the use of high vacuum Sue Carter printed silver paste as thecathode carbon nanotube coating could also be used as the cathode as well as theanode LECs formed by sandwiching the emissive polymer layer between a pair ofcarbon nanotube electrodes were flexible even stretchable if the nanotubes werecoated on elastomeric substrate
The LEC is now intertwined with many other fields It is exciting to witness thelatest progress in LEC performance and exploration of unique applicationsHere I merely state my personal view on what occurred in the past that helped shapethe field as it is today More history and exciting developments are covered by theauthors who wrote the chapters of this book My hatrsquos off to these active researchers
vi Foreword
who have made critical contributions to the field I am indebted to Dr Chi ZhangDr Yang Yang Dr Yong Cao for helping fabricate the first generation of polymerLECs Prof Alan Heeger for polishing the junction model and Dr Gang Yu forsuggesting the planar LEC structure to image the junction
Qibing PeiDepartment of Materials Science and Engineering
Henry Samueli School of Engineering and Applied ScienceUniversity of CaliforniaLos Angeles CA USA
Foreword vii
Preface
The origins of the organic-basedsolid-state lighting (SSL) date back to 1953 but itwas only in the 90s when the organic-based light-emitting diode (OLED) andlight-emitting electrochemical cell (LEC) technologies started to flourish AlthoughOLEDs have made all the way from laboratory to commercial products the LECtechnology is considered as the simplest SSL device The two pillars of LECs arethe type of emitter that holds charge injection charge transport and emission andthe ionic additive that assists charge injection at low applied voltages As intro-duced by Prof Pei in the foreword the LEC revolution is based on the use of ionsto reduce the turn-on voltage After 15 years of research we have gained a matureunderstanding of the device mechanism This has in addition been achieved alongwith the optimization of the two traditional emitters (luminescent conjugatedpolymers and ionic transition metal complexes) the ionic additives for each type ofemitters and the type of poling modes After having fully understood the devicelimitations we have achieved several breakthroughs with respect to the efficiencyusing multilayered architectures (cascade andor tandem) frozen junctions colorconverting layers etc and low-cost and up-scalable fabrication protocols using inaddition unconventional conductive substrates As the most recent research actionwe have focused on investigating different types of emitters like small moleculesnanoparticles quantum dots etc Hence the last two decades have been a suc-cessful test-bed time for LECs reaching both a high industrial relevance and analways-rising research interests as LECs are an easy set-up to investigate theelectroluminescence features of the emitters and the device physics of ionic-basedoptoelectronics
Overall I felt that it was now the right time to bring together all the efforts of theLEC community in this first book devoted to the LEC technology The intention ofthis book is to provide to young students a general description of the LEC tech-nology with a focus on the device mechanism and the different techniques toelucidate the role of mobile anions (Part I) After this general view they will enjoytwo sections specialized on the definition and role of the ionic additives (Part II) aswell as the last advances in traditional and new electroluminescent materials (PartIII) Part II is divided into five chapters that describe in-depth the type of ionic
ix
additives and the different techniques to study the effect of the mobile ions on thedevice mechanism (Chaps 2 to 5) as well as how the ionic electrolytes are crucialfor the fabrication of LECs using deposition tools of industrial relevance (Chap 6)Part III consists of seven chapters summarizing i) the progress in designing iridium(III) complexes (Chap 7) in general and blue-emitting iridium(III) complexes(Chap 8) in particular ii) the studies on new materials with thermally activateddelayed fluorescence features (Chap 9) as well as exciplex emission in conjugatedpolymers (Chap 10) and iii) the last advances in new electroluminescent materialssuch as copper(I) complexes (Chap 11) small-molecules (Chap 12) and quantumdots (Chap 13)
My intention is to provide a comprehensive vision of the past and presentdevelopments in the LEC technology as insights for future advances covering newdevice designs industrial progress and novel types of emitters
Erlangen Germany Rubeacuten D Costa
x Preface
Contents
Part I Introduction to the Light-Emitting ElectrochemicalCell Technology
1 Light-Emitting Electrochemical Cells Mechanismsand Formal Description 3Stephan van Reenen and Martijn Kemerink
Part II Definition and Role of the Ionic Additives
2 Optical-Beam-Induced-Current Imaging of Planar PolymerLight-Emitting Electrochemical Cells 49Faleh AlTal and Jun Gao
3 Optical Engineering of Light-Emitting ElectrochemicalCells Including Microcavity Effect and OutcouplingExtraction Technologies 77Hai-Ching Su
4 The Use of Additives in Ionic Transition Metal ComplexLight-Emitting Electrochemical Cells 93Lyndon D Bastatas and Jason D Slinker
5 Improving Charge Carrier Balance by IncorporatingAdditives in the Active Layer 121Hai-Ching Su
6 Morphology Engineering and Industrial Relevant DeviceProcessing of Light-Emitting Electrochemical Cells 139G Hernandez-Sosa AJ Morfa N Juumlrgensen S Tekogluand J Zimmermann
xi
Part III Traditional and New Electroluminescent Materials
7 Development of Cyclometallated Iridium(III) Complexesfor Light-Emitting Electrochemical Cells 167Catherine E Housecroft and Edwin C Constable
8 Recent Advances on Blue-Emitting Iridium(III) Complexesfor Light-Emitting Electrochemical Cells 203Lei He
9 Thermally Activated Delayed Fluorescence Emittersin Light-Emitting Electrochemical Cells 237Michael Yin Wong and Eli Zysman-Colman
10 White Emission from Exciplex-Based PolymerLight-Emitting Electrochemical Cells 267Yoshinori Nishikitani Suzushi Nishimura and Soichi Uchida
11 Luminescent Cationic Copper(I) Complexes SynthesisPhotophysical Properties and Application in Light-EmittingElectrochemical Cells 287Margaux Elie Sylvain Gaillard and Jean-Luc Renaud
12 Small Molecule-Based Light-Emitting Electrochemical Cells 329Youngson Choe Chozhidakath Damodharan SuneshMadayanad Suresh Subeesh and Kanagaraj Shanmugasundaram
13 Quantum Dot Based Light-Emitting Electrochemical Cells 351Meltem F Ayguumller and Pablo Docampo
xii Contents
Part IIntroduction to the Light-EmittingElectrochemical Cell Technology
Chapter 1Light-Emitting Electrochemical CellsMechanisms and Formal Description
Stephan van Reenen and Martijn Kemerink
Abstract In 1995 Pei and coworkers reported the first light-emitting electro-chemical cell (LEC) LECs are unique because of a simple device layout that doesnot compromise efficient charge injection and transport This is achieved by thepresence of mobile ions in the active layer Despite the simple device layout thedevice physics has proved to be extremely complicated In this chapter an overviewis given of the development and methods towards the current understanding of LECdevice operation A large amount of experimental and modeling work in the last20 years has proved that LECs can operate in different regimes depending oncharge injection and applied bias voltage Processes related to this device operationrange from electric double layer formation at the contacts to electrochemical dopingand the formation of a dynamic p-i-n junction in the bulk of the active layer Wediscuss these and where possible include formal descriptions of transient phe-nomena relating to the turn-on and stability of LECs on the one hand and on theother hand steady-state phenomena relating to the current density potential dis-tribution and recombination properties of LECs
Keywords Light-emitting electrochemical cell Organic light-emitting diode Electroluminescence Organics Scanning kelvin probe microscopy Electricalimpedance spectroscopy
S van ReenenDepartment of Physics University of Oxford Parks road Oxford OX1 3PU UK
M Kemerink (amp)Complex Materials and Devices Department of Physics Chemistry and Biology (IFM)Linkoumlping University SE-58 183 Linkoumlping Swedene-mail martijnkemerinkliuse
copy Springer International Publishing AG 2017RD Costa (ed) Light-Emitting Electrochemical CellsDOI 101007978-3-319-58613-7_1
3
11 Purpose and Aims
The light-emitting electrochemical cell (LEC) is a subclass of organic optoelec-tronic devices which generates light from electricity by means of electronndashholerecombination in an organic semiconductor The major characteristic of LECs is theinclusion of mobile ions in the active layer which leads to complex device physicsand more importantly efficient and robust operation in a simple device configu-ration As such various material classes can be used to produce LECs This chapterexpands on the general description of LECs with a focus on development towards acommon universal understanding of the LEC device operation Furthermore weprovide a formal description of this operation that enables numerical and analyticalmodels to qualitatively and quantitatively describe LEC device behavior
Section 12 will provide the reader with a general overview of LECs in terms of(i) the discovery and development of LECs (ii) a classification in three differentclasses (iii) the device architecture and figures-of-merit (iv) the formulation of theoperational model The main intention is to provide the reader with a precisebackground to assist the understanding of the following chapters Sections 13 and14 will then concentrate on the study of respectively transient and steady-stateprocesses in LECs The final Sect 15 summarizes the current understanding ofLECs and provides an outlook on future challenges and work regarding theunderstanding of LECs
12 Overview
The discovery of LECs in 1995 was preceded by various other discoveries relatingto lighting from organic semiconductors Here a short overview is given on keyevents preceding the establishment of LECs as a class of organic lighting devicesFurthermore we will introduce the classification of LECs into three main classesbased on the type of semiconductor ie conjugated polymers (CP) ionic transitionmetal complexes (iTMC) and the so-called third generation spanning smallmolecules quantum dots and luminescent nanoparticles to be discussed in moredetail in Sect 3 Finally we discuss in depth the suggested operational mechanismsof LECs and give an overview of the current model of LEC device operation
121 Background
Electroluminescence from organic materials was first reported by A Bernanose andcoworkers in 1953 when they applied high alternating voltages to films based oneg acridine orange [1] Subsequent development of organic electroluminescent
4 S van Reenen and M Kemerink
devices took however until 1987 due to the requirement of high drive voltages andconsequently poor power efficiencies
In 1987 Tang and VanSlyke reported the first organic light-emitting diode(OLED) which started the current period of global OLED research and develop-ment [2] They produced a diode with a bilayer structure by vapor deposition of thinfilms of organic semiconductors Electrodes were used with such work functionsthat carrier injection in the organics was relatively efficient Furthermore the bilayerstructure enforced recombination at the interface between the two organic semi-conducting films which avoided significant losses related to electroluminescencequenching at the electrodes
A few years later in 1990 Burroughes et al introduced an OLED based on CP[3] One advantage of such polymers is that they enable a larger range of depositiontechniques as they can be processed from solution to form uniform thin filmsFurthermore due to their macrostructure they can in principle lead to higherlong-term stability by avoiding eg recrystallization or other structural changeswhich can be expected for films based on small molecules
For efficient OLED operation electrons and holes need to recombine in the bulkof the organic semiconductor to avoid electroluminescence quenching at the elec-trodes To achieve this efficient and balanced carrier injection and transport in theorganics is paramount Electron injection in organic semiconductors occurs in thelowest unoccupied molecular orbital (LUMO) whereas hole injection occurs in thehighest occupied molecular orbital (HOMO) In efficient OLEDs one high- and onelow-work function electrode are therefore required for efficient hole and electroninjection respectively This leads to the typical diode-like behavior of these devices[4] Injected charge carriers form positively and negatively charged polarons thatmove through the organic semiconductor with a given mobility that can vary fordifferent types of semiconductor [5] Carrier transport through the organic occursthrough thermally activated tunneling or lsquohoppingrsquo for short of the charge carriersbetween localized sites in a disordered energy landscape [6 7] As mobilities arerelatively low in organic semiconductors used for OLED applications the organicfilms in OLEDs typically need to be extremely thin to achieve efficient operation atstill acceptable biases in the range of hundreds of nanometers Electroluminescencein OLEDs occurs via recombination of electrons and holes that form singlet andtriplet exciton pairs that can decay radiatively decay non-radiatively or dissociateThe emitted phonon typically has an energy that is close to the bandgap of theorganic semiconductor One major source of non-radiative decay is excitonquenching near the metallic injecting electrodes [8ndash10] This is likely to occur incase of a large imbalance in carrier injection andor bulk carrier transport propertiesbetween electrons and holes In OLEDs issues like these are typically avoided byadopting a bilayer structure [2] or by introduction of additional organic layers [11]that facilitate injection and transport of one polarity of charge carrier whileblocking the other In modern OLEDs such layers are used on both sides of theemissive layer(s) which leads to a massive increase in device complexity andfabrication cost [12]
1 Light-Emitting Electrochemical Cells hellip 5
During the development of the OLED a new type of organic light-emittingdevice was discovered by Qibing Pei and coworkers in 1995 [13] By includingmobile ions in the organic semiconductor they were able to produce a device withstrongly contrasting properties compared to the OLED The device had a simpledevice layout consisting of only one mixed organic layer sandwiched betweenmetallic electrodes as shown in Fig 11a Furthermore the device did not behavelike a diode as it did not block current in reverse bias despite the asymmetry createdby the use of one high- and one (moderately) low work-function electrode Thisbehavior is illustrated in the current-voltage characteristic shown in Fig 11bRemarkably the light intensity is similar for both bias polarities showing that thedevice is as efficient in forward as in reverse bias conditions despite the presence oflarge injection barriers for at least one of the biasing conditions This highlights thatdespite large injection barriers this new type of device which was coined LEC canoperate efficiently at low voltages as opposed to OLEDs The turn-on voltage of thecell is observed to be around the bandgap of the semiconductor Another strikingproperty of LECs is shown in Fig 11c Efficient LECs can be produced in a planarconfiguration with an interelectrode spacing of in this case 15 lm between goldelectrodes At a bias voltage of 4 V light emission can be observed which showsthat operation at low voltage is still possible despite a thick or in this case wideactive layer Moreover it indicates that efficient operation is possible while usingair stable materials to contact the active layer despite the consequent large injectionbarrier for electron injection
The active layer of the first LECs shown in Fig 11 consists of a CP that ismixed with a salt and an ion-solvating polymer poly(ethylene oxide) (PEO)However the combined use of ions and organic semiconductors in LECs is notlimited to these materials For instance iTMCs are luminescent ionic materials thatconsist of a metal center surrounded by organic ligands The first electrolumines-cent device based on iTMCs was published by Lee et al in 1996 [15] They used atris(bathophenanthrolinedisulfonate)ruthenium(II) sodium salt with a bandgap of
Fig 11 Left Schematic device layout of a CP-based LEC including the molecular structures ofthe active layer contents Central typical current-voltage-luminescence characteristic of aCP-based LEC (redrawn after Ref [13]) Right photograph of a planar CP-based LEC(reproduced from Ref [14])
6 S van Reenen and M Kemerink
roughly 26 eV as shown in Fig 12a This device was not reported as an LEC andonly a current-voltage-luminance (IVL) characteristic in forward bias conditionswas reported as shown in Fig 12b It is however interesting that the spin-coateddevice (circles in Fig 12b) despite having Al and ITO electrodes that would giverise to substantial injection barriers shows a turn-on in current and luminescencefor voltages near the bandgap which suggests LEC-like operation Later reports in1999ndash2001 showed slow time components in the electrical characterization insimilar cells which are attributed to slow ionic transport within the active layer [16ndash18]
Recently the scientific community started with the development of the so-callednext generation of electroluminescent materials for LECs [19ndash32] Small moleculesfeaturing thermally delayed activated fluorescence sustainable copper(I) com-plexes quantum dots and luminescent nanoparticles are leading examples of thisnew type of LECs see section III of the book Noteworthy these scattered worksindicate that the LEC behavior holds independently of the type of materials andadditives used As such the current chapter will focus on providing a generaldescription of LEC operation that is predominantly based on works focusing on CP-and iTMC-based LECs
122 Figures of Merit and Device Architectures
Possible applications of LECs are commonly thought to be found in domains wherethe benefits associated with facile large-area fabrication are maximal like lightingsignage and possibly displays Consequently LEC performance is typicallydescribed by the following figures of merit
1 The turn-on time (ton) which is defined by the time between switch-on of thecell by application of a bias voltage and the time at which the luminance reaches
Fig 12 Left Schematic device layout of an iTMC-based LEC including the molecular structureof the active layer contents Right current (open symbols) and luminance (closed symbols)characteristics of iTMC-based LECs in which the iTMC layer is formed by spin-coating (circles)and self-assembly (triangles) (reproduced from Ref [15])
1 Light-Emitting Electrochemical Cells hellip 7
a certain predefined level typically equal or close to the quasi-steady-stateluminance
2 The luminance (cdm2) describes the amount of luminous power per unit areacorrected for the wavelength-dependence of the sensitivity of the human eye Itgives a measure of the brightness of the LEC (Bmax)
3 The electroluminescent efficiency (Effmax) expressed either in lmW (powerefficiency) or cdA (efficacy) where 1 lm = 1 cdsr This efficiency relates to theconversion efficiency of electronic carriers in photons weighted by the sensi-tivity curve of the human eye Note that when expressing the efficiency in lmWunlike cdA the bias voltage that drives operation is taken into account
4 External quantum efficiency (EQE) The ratio of photons emerging from thedevice per injected electrons EQE is also defined through the equationEQE=b2n2 where b is the recombination efficiency (equal to unity for twoohmic contacts) is the fraction of excitons that decay radiatively (photolu-minescence quantum yields) and n is the refractive index of the glass substrateand is equal to 15 (the factor 12n2 accounts for the light outcoupling of thedevice)
5 The lifetime (t12) which can be defined either on-shelf or under operation Forthe latter one can take as a measure eg the time for the luminance to decay tohalf-maximum or to below a certain threshold that can depend on the foreseenapplication Another parameter is the total emitted energy (Etot) It is calculatedby integrating the radiant flux of the device versus time from t=0 (application ofbias) to t=t15 If this value is divided by the electrode area it yields the totalemitted energy density Utot which allows devices having electrodes of differentshapes to be compared
6 Color coordinates are used to compare color in a standardized manner Thehuman eye has color sensors in three different wavelength regions short (max440 nm) medium (max 535 nm) and long (max 567 nm) which allow todifferentiate about 10 million colors The number is actually low compared tothe amount of colors defined by the Commission Internationale de lrsquoEclairage(CIE) in 1931 It states that every color can be described by the three colormapping functions x(k) y(k) and z(k)
For research purposes typically two LEC architectures are used stacked ordiode-type illustrated in Fig 13a and planar illustrated in panel b of the samefigure The major advantage of the latter is the good access to the (various parts inthe) active layer by microscopic techniques see eg Figs 11 16 and 110 thesmall relative area of the stripe-like emissive zone in planar LECs (Fig 11c) makesthis geometry effectively useless for practical applications As compared to themultilayer stacks encountered in modern OLEDs the stacked LEC layout isextremely simple In combination with the inherent tolerance to layer thicknessvariations which follows from the operational mechanism that is discussed in thenext sections this rationalizes the main asset of LECs low cost by virtue of facilesolution-based fabrication It should however be kept in mind that the total
8 S van Reenen and M Kemerink
manufacturing costs of an LEC module also include materials costs for the activelayer (synthesis) as well as for the substrate electrodes and packaging
Regarding future application of LECs it is important to determine the keyfeatures of LECs as opposed to related technologies ie OLEDs and PLEDsTable 11 gives a rough overview and summarizes the benefits of LECs Apart fromadvantages on conventional substrates like foil or glass LECs can also be easilyfabricated on a large range of different substrates like fibers that can possibly beused to produce light-emitting clothing [33] paper [34] and even oncomplex-shaped surfaces like kitchen forks [35] Current industry interest seemsfocused on OLEDs because of their energy efficiency color quality andhigh-contrast compared to LCD technology A key selling point of organicshowever has always been to become a low-cost technology which to date has not
Fig 13 Schematic layout of a stacked and planar LECs
Table 11 Comparison of different types of organic light-emitting devices PLEDs OLEDs andLECs
Parameter PLEDs OLEDs LECs Benefits of LECs
Active layers 1ndash3 4 or more 1 or 2 Simple devicearchitecture
Typicalthickness perlayer
60ndash120 nm 1ndash40 nm 100ndash500 nm Thicker films promiserobust processes
Cathode Air sensitive Air sensitive Air stable Air stable metals like AlAg Au can be used
Encapsulationrequirements(permeationrate of H2O)
High (10minus6
gm2day)High (10minus6
gm2day)Low(10minus6
gm2day)
Air stable electrodespromise less demandingpackaging
Processing oforganic layers
Solution-based Vacuum-based Solution-based Cost-efficient R2Rprocessing
Solvent Aromatic na Benign(alcoholswater etc)
Environmental friendly
1 Light-Emitting Electrochemical Cells hellip 9
been achieved LECs however can make a strong case to fill this gap as recentlydescribed by Sandstroumlm and Edman [36] If LECs are produced with a reasonableluminance of 1000 cdm2 in a high-volume roll-to-roll-coating scenario then thecost per lumen would be roughly 00036 eurolm This cost would be one order ofmagnitude lower than the projected future costs for LEDs and OLEDs
123 Suggested Operational Mechanisms for LECs
From a processing and fabrication perspective the single-layer architecture of LECsis a major benefit However the mixing of ionic and electronic conductors has ledto a significant complexity of the device physics Understanding the device physicsfurthermore suffered hereof furthermore suffered from the large variety of materialsstudied in LECs on top of variations that may be expected to arise between worksdone in different laboratories With present knowledge it is quite understandablethat the fact that LECs can exhibit rather different behaviors depending on subtleparameter variations has led to confusion Over time various operational mecha-nisms have been suggested in the literature which are summarized below
1231 Electrochemical Doping Model (ECDM)
The electrochemical doping model was first proposed by Pei et al [13 37] and lateron supported by theoretical work by Smith [38] and Manzanares et al [39]According to this model the LEC operation is mainly determined by electro-chemical doping of the active layer that results in the in situ formation of a p-i-nstructure A schematic of this model is shown in Fig 14a Injection of electronsand holes leads respectively to the oxidation and reduction of the semiconductor atthe anode and the cathode The oxidized and reduced semiconductor is electro-statically compensated by anions and cations respectively which results in p-typeand n-type doped regions in the bulk The region between the doped regionsremains intrinsic and is the place where the injected electronic charges recombineA relatively large electric field is present in the intrinsic region which compensatesits relatively low conductivity This electric field is due to space charge in thejunction that originates from electronic carriers This is in contrast to conventionalp-n junctions in which the space charge in the junction originates from dopant ions[38] The doped regions at the electrode interfaces lead to low-resistance contactsthat allow efficient carrier injection despite large injection barriers
1232 Electrodynamic Model (EDM)
The electrodynamic model was first proposed by deMello et al [41ndash43] Accordingto this model the LEC operation is mainly determined by the formation of electric
10 S van Reenen and M Kemerink
double layers at the interfaces as shown in the schematic in Fig 14b The electricdouble layers are formed by mobile ions in the active layer that upon application ofa field drift towards the electrode interfaces and create large interfacial electricfields The ions continue to move towards the interfaces until the bulk of the activelayer is field-free The electric fields created at the interfaces promote the injectionof electronic charge carriers The injected carriers then move through the bulk bydiffusion and recombine when electrons and holes meet
1233 Preferential Electrochemical Doping Model (PECDM)
The preferential electrochemical doping model was first proposed by Leger et al[44] This model shows high similarity to the ECDM with the main difference thatonly one type of doping occurs in the active layer either n-type or p-type doping[44ndash46] As a result recombination occurs close to one of the electrode interfaces
Fig 14 Schematics of the ECDM (left) and the EDM (right) with the associated spatialdistribution of the electric field shown below (reproduced from Ref [40])
1 Light-Emitting Electrochemical Cells hellip 11
124 Current Understanding of Operational Mechanismof LECs
The variation in operational behavior of LECs reported experimentally andnumerically indicates that LECs operate in different regimes Systematic experi-ments [46 47] and numerical modeling [47 48] have confirmed this fact andshowed that the three models the EDM the PECDM and the ECDM coexist Theapplicability of each model was found to depend on the ability to form Ohmicinjecting contacts [47] which depends on a combination of applied bias voltage andthe height of the barriers for carrier injection [47 48] In case no Ohmic contacts areformed the LEC follows the EDM In case one Ohmic contact is formed the LECfollows the PECDM Here p-type doping or n-type doping occurs in case an Ohmiccontact is formed at the anode or the cathode respectively In case two Ohmiccontacts are formed the LEC follows the ECDM An overview of this unifyingmodel is shown in Table 12
This unifying model has been confirmed in both CP- and iTMC-based LECs byvarious experimental and numerical studies [13 14 38 41 45ndash47 49ndash55] Froman application perspective device operation in the ECDM mode is preferred [47]
125 Basic Equations to Describe LEC Operation
Here the basic equations are summarized that describe the various processes relatedto charge injection transport and recombination in the active layer of LECsLateral heterogeneity that can arise from for example phase separation will be
Table 12 Overview of the universal operational mechanism of light-emitting electrochemicalcells
Light-emitting electrochemical cells
Requirement (dependson applied bias voltageand injection barriers)
0 ohmic contacts 1 ohmic contact 2 ohmic contacts
Operating mechanism Electrodynamicmodel (EDM)
preferentialElectrochecmicaldoping model(pECDM)
Electrochecmicaldoping model(ECDM)
Voltage distribution(back line) andrecombination zone(orange)
+ -Can be anywhere in active layer
+ - + -
Electron (bull) Hole(∘)anion (oplus) cation (⊖)distributions
+ -+++
---
+ -++
---
++
+
++++ + + -+
+++
+
++--
---
12 S van Reenen and M Kemerink
ignored Hence a 1D model suffices Although computationally cumbersomeextension to 2D or 3D is straightforward
1251 Drift and Diffusion for Ionic and Electronic Charges
Transport of electrons holes anions and cations can be described by drift anddiffusion where the diffusivity D is related to the mobility l by the Einstein relationDi frac14 likBT
Ji frac14 niqlidVdx
thorn bDidnidx
eth11THORN
where b is +1 for electrons (i = n) and anions (i = a) and minus1 for holes (i = p) andcations (i = c) Moreover Ji is the current density of species i ni the charge densityq the elementary charge V the electrostatic potential x the position kB theBoltzmann constant and T the temperature
1252 Poissonrsquos Equation
The electrostatic potential and field and the density of electrical charge carriers arerelated through Poissonrsquos equation
r2V frac14 rE frac14 qe0er
np nn thorn nc na eth12THORN
where E is the electrostatic field e0 the vacuum permittivity and er the relativepermittivity of the material
1253 Binding Energy for AnionCation and IonElectronic ChargePairs
The mobile ions in the active layer of LECs can recombine with each other to formneutral salt molecules with a binding energy Eac The anionndashcation capture rate Cacdissociation rate Bac and net binding rate Uac can be described respectively by
Cac frac14 cac na nc eth13THORN
Bac frac14 Kac bac nac eth14THORN
Uac frac14 Cac Bac eth15THORN
1 Light-Emitting Electrochemical Cells hellip 13
where cac is the anionndashcation capture rate coefficient and nac the density of boundanionndashcation pairs Kac is a mass-action law constant and bac the anionndashcationdissociation rate coefficient that can be described by
bac frac14 cac exp Eac
kT
eth16THORN
Although this has not been explored a binding energy is also expected for dopingcomplexes ie between anions and holes (Eap) and between cations and electrons(Ecn) In analogy to Eqs (13)ndash(16) the anion-hole capture rate Cap dissociationrate Bap and net binding rate Uap can be described respectively by
Cap frac14 cap na np eth17THORN
Bap frac14 Kap bap nap eth18THORN
Uap frac14 Cap Bap eth19THORN
where cap is the anion-hole capture rate coefficient and nap the p-type dopingdensity Kap is a mass-action law constant and bap the anion-hole dissociation ratecoefficient that can be described by
bap frac14 cap exp Eap
kT
eth110THORN
Cation-electron binding can be described by expressions similar to Eq (17)ndash(110)
1254 ElectronndashHole Recombination
Recombination between electrons and holes in electroluminescent materials leads tothe formation of exciton complexes that can decay radiatively The materials usedin LECs are typically low dielectric constant materials leading to strong Coulombicinteraction between the electron and hole in the exciton complex As a result theexcitons are relatively small and typically reside on a single unit cell
The recombination process can be described by either Langevin recombination[56] or trap-assisted recombination ie Shockley-Read-Hall (SRH) recombination[57 58] In case of Langevin recombination the recombination occurs between freeelectrons and holes that drift towards each other In case of SRH recombinationrecombination occurs between a trapped and a free carrier Here the recombinationrate is set by the mobility of the free charge carrier Hence both Langevinrecombination and SRH recombination are governed by charge carrier mobilities[59] The Langevin recombination rate RL is described by
14 S van Reenen and M Kemerink
RL frac14q ln thorn lp e0er
npnn eth111THORN
1255 Continuity Equations
The continuity equations for electrons holes anions and cations can be describedrespectively by
dnndt
frac14 1qdJndx
RL Ucn eth112THORN
dnpdt
frac14 1qdJpdx
RL Uap eth113THORN
dnadt
frac14 1qdJadx
Uac Uap eth114THORN
dncdt
frac14 1qdJcdx
Uac Ucn eth115THORN
Continuity equations for salt and doping complexes can be described respec-tively by
dnsdt
frac14 Uac eth116THORN
dnapdt
frac14 Uap eth117THORN
dncndt
frac14 Ucn eth118THORN
1256 Boundary Conditions
The density of the electronic charge carriers at the interfaces are determined by themodel that describes carrier injection A realistic model is the Emtage-OrsquoDwyermodel [60] that describes injection with an exponential electric field dependenceThis model also considers the injection barrier created by the energy level offsetbetween the electrode Fermi-level and the transport level in the semiconductor inwhich the charge carriers are injected In addition it also considers stabilization ofinjected charge carriers in the semiconductor by the formation of an image charge inthe electrode Implementation of the Emtage-OrsquoDwyer injection model is numeri-cally challenging as a relatively small grid-point spacing is required which can lead
1 Light-Emitting Electrochemical Cells hellip 15
to an enormous grid when equal grid-spacings are required [61] deMello [41]solved this problem by using an adaptive grid with a variable grid-point spacingVan Reenen et al [47] used a lsquomodified Boltzmannrsquo model which ignores thegrid-point spacing related to carrier injection but simply assumes thermal activationof charge carriers over an injection barrier that is modified by the local electric fieldIn LECs this field results typically from the formation of an electric double layer atthe interfaces due to pile-up of ions
For ions the electrodes are typically assumed blocking Experiments in planarcells have however suggested that ions can penetrate and even travel through Auelectrodes [14]
13 Transient Phenomena
131 Turn-on and the Role of Ion Motion
The presence of ions in LECs makes operation of these cells stronglytime-dependent Processes related to electronic transport in organic semiconductorstypically take far less than a second to reach quasi-steady state Ions however aremuch slower as they physically move through a solid-state material leading toturn-on times that are relatively long as compared to OLEDs Turn-on times inLECs can range from several seconds [13] to several hours [62] dependent on thecombination of the constituents and the active layer thickness as well as the appliedbias voltage
For the vast amount of LEC device configurations studied and reported in theliterature large quantitative differences are observed in the turn-on transientsNevertheless the turn-on transients of CP-based LECs in stacked [63 64] andplanar [47 65ndash68] configuration show strong qualitative resemblance withiTMC-based LECs in stacked [18 54 62 69ndash72] and planar configuration [53]This resemblance was studied systematically by Van Reenen et al [55] Theystudied the time-dependent current luminance and efficacy just after switch-on offreshly prepared CP- and iTMC-based LECs They found that the timescale atwhich turn-on occurs is strongly affected by the device temperature By normalizingto the turn-on time this temperature dependence is scaled out Furthermore if thecurrent luminance and efficacy are normalized as well the transients of both typesof LECs are found to follow a universal shape as shown in Fig 15 Moreover theactivation energy of the turn-on time and the ion conductivity measured in theoff-state were found to be the same which substantiates that the turn-on of LECs isdetermined by the ion conduction
In the remainder of this paragraph the turn-on behavior of planar and stackedLECs is reviewed separately in more detail Being not fundamental the differencesin turn-on between CP- and iTMC-based LECs will not be considered
16 S van Reenen and M Kemerink
1311 Studies in Planar LECs
The ability to construct functional LECs in planar cell configuration has givenresearchers access to the active layer for various experimental surface techniquesAs shown in Fig 11c this allowed researchers to study the electroluminescence(EL) position in LECs Besides EL also photoluminescence (PL) can be studied inplanar LECs using UV illumination to excite the semiconductor Electrochemicaldoping of semiconductors is known to quench PL [73] Hence this techniqueallows one to keep track of the electrochemical doping process in LECs duringturn-on [50 53 66 74] A typical example of such UV-excited PL measurements inplanar LECs is shown in Fig 16a [50] The experiments show that p- and n-typeelectrochemical doping occurs in the LEC by accelerating doping fronts that movethrough the active layer starting from the electrodes until both meet [75] Theposition where the p- and n-type doping fronts meet coincides with the regionwhere light emission takes place as shown in the last photograph in Fig 16a Thecorresponding device current grows during front propagation and continues to growafter the fronts connect (Fig 16b)This continuous growth in current is attributed toon-going electrochemical doping of the doped regions The enhanced quenching ofPL observed in Fig 16a after doping front connection further constitutes this viewOther work has also shown that besides this continuation of doping after frontconnection the recombination zone can also shift towards the anode or the cathodewith time [53 76]
Van Reenen et al studied the turn-on in planar cells by means of a numericaldrift-diffusion model based on the equations described in Sect 125 It was pos-sible to calculate a current transient (see Fig 16c) that qualitatively reproduces theexperiment (Fig 16b) [77] The potential profile evolution in the active layer wasstudied experimentally by use of scanning Kelvin probe microscopy (seeFig 16d) and numerically (see Fig 16e) This combined work gives the fol-lowing picture of the turn-on of planar LECs
After switch-on of the bias voltage electric double layers form at the interfacessee Fig 16d e enabling carrier injection and consequently n-type and p-type
Fig 15 Normalized currentluminance and efficacytransients of CP- andiTMC-based LECs at twotemperatures each and biasedat 35 V (reproduced fromRef [55])
1 Light-Emitting Electrochemical Cells hellip 17
electrochemical doping of the active layer The electrochemically doped regionsgrow towards the opposite electrodes until they meet When the fronts connect thepotential is distributed more or less evenly across the active layer as shown inFig 16d (at t 10 s) and e (as indicated) Doping continues as there are stillmobile ions available that do not yet contribute to electrochemical doping Hencethe current continues to grow The current reaches a peak value in both the modeland experiment Around this time the potential is observed to change dramaticallythe potential becomes distributed mainly in the region where recombination takesplace as shown in Fig 16d (at t 30 s) and e (as indicated) The model showsthat at the same time the recombination zone becomes depleted of ions so that the
Fig 16 Top Photographs under UV illumination of planar CP-based LECs with an electrode gapof 90 lm during operation at V = 8 V and T = 333 K The positive and negative electrodes areindicated in the photographs and the dashed lines indicate where they contact the active layerCentral left Experimental current measured during turn-on of the planar CP-based LEC shown atthe top pictures Central right Modeled current in a planar CP-based LEC Bottom leftExperimental potential profile evolution during turn-on of a similar planar CP-based LEC asshown at the top pictures Bottom right Modeled potential profile evolution of a planar CP-basedLEC (adapted from Ref [77])
18 S van Reenen and M Kemerink
recombination zone becomes intrinsic ie undoped and therefore has a lowerconductivity This lower conductivity in the recombination zone as compared tothe high conductivity in the doped regions necessitates a larger field in therecombination zone to have current conservation across the device The voltageredistribution towards the recombination zone leads to reduction in current in thelater stages of the transients (see Fig 16b c)
The formation of doping fronts in planar LECs to electrochemically dope theactive layer has been a major topic of study as it reveals various interestingproperties regarding carrier mobilities and local field distributions
A report by Shin et al showed that the doping fronts typically accelerate duringpropagation as shown in Fig 17a (blue circles) [65] Robinson et al used ananalytical model to describe this behavior The model is based on matching thecurrent density through the doped regions with the ion current in the intrinsicregion They assumed that the applied potential drops nearly completely over theintrinsic region during the turn-on of the cell because of the mismatch in con-ductivity between the intrinsic region and the doped regions A fit of the analyticalmodel shown in Fig 17a (red line) shows that the model indeed predicts accel-erating doping fronts The assumption regarding the potential distribution however
Fig 17 Top left Normalized front position during switch-on in an LEC as obtainedexperimentally (blue circles) [65] analytically (red line) [78] and numerically (black line) [79]Top right Schematic of the ion mobility criteria with respect to the doping-dependent electron andhole mobility (blue line) that result in the formation of accelerating doping fronts in LECs duringswitch-on Bottom Transient electron (closed circles) and hole (open circles) density profiles inLECs for a constant (left) and doping-dependent electronhole mobility (right) and an ion mobilityof 510minus11 m2Vminus1sminus1 The anode and cathode are positioned at position = 0 and 2000 nmrespectively Movement in time is expressed by the color change of the graphs from light gray toblack to red (reproduced from Ref [79])
1 Light-Emitting Electrochemical Cells hellip 19
was later proven to not necessarily be true [77] as shown in Fig 16d e thepotential profile hardly changes during doping front propagation
An attempt to model doping front propagation numerically was made using adrift-diffusion model of a planar LEC [79] It was found that doping-dependentelectron and hole mobilities are required to achieve pronounced doping fronts(Fig 17d) Without such a doping dependence electrochemical doping of theactive layer would occur more gradually as shown in Fig 17c More specificallyto achieve accelerating doping fronts the mobility of electrons and holes in theundoped region lpn0 needs to be similar in order of magnitude to the ion mobility(schematic in Fig 17b) The doping density dependence of the mobility in elec-trochemically doped systems is relatively strong compared to eg field effectmobility enhancements [80] A low mobility in weakly doped semiconductors islikely due to localization of the electronic carriers at the electrostatically compen-sating doping sites [81] doping sites act like charge traps at low doping densitiesAt higher doping densities this localization is lifted as there are more neighboringdoping sites available to hop to without the need to gain additional energy to escapethe energetically favorable trap [80 81] Hence the mobility of the charge carrier issignificantly enhanced
Another typical feature of doping fronts in planar LECs is the instability at thedoping fronts which leads to the formation of lsquofingersrsquo An example of this isshown in Fig 18a This instability was furthermore found to be enhanced by theuse of higher operational voltages [78] To understand this behavior Bychkov et alcarried out various numerical modeling studies on doping front propagation [8283] They found that the local electric field at the apexes of the doping front isrelatively high (see Fig 18b c) This increase of the local electric field thenaccelerates the doping front propagation locally which leads to an enhancement ofthe finger shape This process explains the instability at the doping fronts as shown
Fig 18 Left top Experimental photo of a p-type doping front in a planar CP-based LEC Thewhite line is a simulation of the p-type front shape Left central and bottom show the relativeincrease in the electric field in the undoped region obtained numerically for b the whole front andc a selected part with equal color scales as indicated below Right Schematic of the p- and n-typedoping fronts in a LEC (reproduced from Ref [82])
20 S van Reenen and M Kemerink
in Fig 18a and allowed the researchers to model a similarly shaped doping front(see Fig 18a white line) [82]
1312 Studies in Stacked LECs
Like in planar cells the turn-on time in stacked cells is dominated by ion transportA logarithmic plot of a current transient determined from a CP-based LEC isshown in Fig 19a Van Reenen et al carried out a study to model these transients[63] By experimental determination of the ion conductivity using electrochemicalimpedance spectroscopy [63] they tried to use the determined ion mobility anddensity to model the turn-on transient of the same CP-based LEC see Fig 19dashed line The timescale of the modeled transient was however found to becompletely off compared to the experiment eg the modeled current reached aquasi-steady-state around t = 100 s whereas in the experiment it took over 101 s todo so Only by including a binding energy (see Eq (13)ndash(16)) between the ions inthis case of 015 eV [38] transients could be modeled that took over 100 s to reachsteady-state while taking the measured ion conductivity values into account (seeFig 19b solid line)
Turn-on times in the order of seconds or even larger are not suitable for variousapplications of LECs like displays Consequently much effort has been put intoimprovement of the turn-on time Addition of the ion-dissolving poly(ethyleneoxide) (PEO) has been found to accelerate turn-on significantly in CP-based LECs[13] as well as iTMC-LECs [69] This addition possibly leads to improved ionconduction and enhanced salt dissociation as PEO has a relatively high dielectricconstant of 6 as compared to the dielectric constant of typical CPs around 3 [84]Improvement of the electrolyte by use of different ion-solvating polymers hasalready led to significant improvement of the turn-on time in LECs besidesimproved lifetime [85ndash88] Ion conduction and salt dissociation also depend on the
Fig 19 Left Experimental current transient of a pristine CP-based LEC biased at 35 V (adaptedfrom Ref [55]) Right Modeled current transient of a LEC biased at 35 V with (straight line) andwithout (dashed line) binding energy between anions and cations (reproduced from Ref [63])
1 Light-Emitting Electrochemical Cells hellip 21
type of ions used in LECs Therefore the use of other ions [65 89 90] or ionicliquids [91] in LECs can also lead to improved turn-on times
Turn-on can also be accelerated by driving LECs with a fixed current instead ofvoltage [62 92 93] This typically translates into a relatively high initial biasvoltage applied to the cell when it is in its least conductive state Use of a pulseddriving scheme was furthermore shown to be able to stabilize the device in anintermediate state during turn-on of the cell preventing complete electrochemicaldoping of the active layer and the associated efficiency roll-off that is furtherdiscussed in Sect 144 [94]
Another method to avoid long turn-on times in LECs is by preparing the LEC insuch a way that electrochemical doping is fixed This can be done chemically byfixing doping complexes using eg polymerization reactions of dopant ions [9596] Alternatively ions can be frozen into position by lowering the device tem-perature after electrochemical doping [97ndash99] This topic will be discussed furtherin the next section
132 Polarization Reversal and Hysteresis
The dynamic character of mobile ions gives LECs the unique feature that they canbe operated efficiently both in forward and reverse bias conditions as illustratedalready in Fig 11b Polarization reversal was studied in planar cells by scanningKelvin probe microscopy by Matyba et al [14] Steady-state profiles of thepotential are shown in Fig 110a and c where a planar CP-based LEC was sub-sequently biased +5 V and minus5 V The potential profiles are essentially mirrorimages of each other which indicates that the processes related to the operation ofLECs are highly reversible and therefore enable polarization reversal without sig-nificant penalties relating to irreversible side reactions Figure 110b shows thecomplex evolution of the potential profile during a polarization reversal
Despite the electrochemical doping process being essentially reversible hys-teresis effects are significant in LECs Li et al carried out various experiments tostudy the effects of relaxation of doping in stacked CP-based LECs [100]Prolonged operation of LECs typically leads to reduction of the electrolumines-cence by electrochemical doping similar to PL quenching in planar cells (see egFigure 16a) Li et al studied the recovery of luminance by relaxing the cell atopen-circuit voltage after prolonged operation as shown in Fig 111 The cellsrequired a relatively long time at elevated temperatures (to speed up the relaxationprocess) to recover a large part of the luminance During this time part of theelectrochemical doping is removed by dedoping of the cell The PL images inFig 111 (compare top and middle photograph of the device) furthermore confirmthat PL quenching is significantly reduced after allowing the cell to relax for 7 h at60 degC Although part of the luminance is recovered after a long time of relaxationanother part did not recover which indicates the occurrence of irreversible reactionsin the LEC during operation These will be further discussed in Sect 133
22 S van Reenen and M Kemerink
Reduction of hysteresis in LECs can be achieved by fixation of the electro-chemical doping One method is to pre-bias the device at elevated temperature atwhich ions are mobile followed by reduction of the device temperature leading tosignificant reduction of the ion conductivity [97ndash99] For practical applications thelatter requires that the freeze-in temperature is significantly above room tempera-ture Edman substituted PEO in planar CP-based LECs by a crown ether that meltsat a temperature of 56 degC [98] The device was switched on at 85 degC and subse-quently cooled down to room temperature Consequently the p- and n-type dopedregions were frozen-in by crystallization of the crown-ether phase The same can beachieved when using PEO by cooling the device down to 100 K which is belowthe glass transition temperature of PEO [97] In both instances the response timesof the LECs improved significantly enabling frozen-junction LECs to match theresponse times of similar polymer-based LEDs [97]
Chemical binding of doping offers an alternative solution to reduce hysteresisand improve turn-on of LECs [95 96] Hoven et al reported an LEC that consistedof a bilayer structure [95] This bilayer comprised a film based on a cationicconjugated polyelectrolyte with mobile fluoride counter (an)ions (PFP-F) and a filmbased on a neutral CPs with functional groups that enabled trapping of fluorideanions (PFP-BMes) as shown in Fig 112 By application of a bias voltage themobile fluoride anions move from the PFP-F film into the PFP-BMes film to enableelectrochemical doping The fluoride is then covalently bonded to the PFP-BMesAs a result these devices have reduced hysteresis and do not relax back to the
Fig 110 Experimental dataillustrating the polarizationreversal in planar CP-basedLECs by scanning Kelvinprobe microscopySteady-state operation atVbias = +5 V (top) Temporalevolution after subsequentswitch to Vbias = minus5 V wherethe arrow indicates the time(central) Steady-stateoperation at Vbias = minus5 V(bottom) The dashed linesindicate the electrodepositions as determined fromheight data from atomic forcemicroscopy measurements(reproduced from Ref [14])
1 Light-Emitting Electrochemical Cells hellip 23
pristine state by dedoping after initial charging Furthermore this device showedrectification unlike typical dynamic LECs
133 Degradation Side Reactions and ElectrochemicalStability
Before discussing degradation and side reactions in LECs it is important to notethat reduction in luminance and efficiency as well as current over time in LECs isnot necessarily related to irreversible side reactions Luminescence quenching and areduction in efficiency are generally observed in LECs but can be partially recov-ered by relaxation of the cell as illustrated in Sect 132 Meier et al studiedreversible and irreversible effects in stacked iTMC-LECs during operation bymeasurement of the PL intensity as a function of operating time (see Fig 113) [52]At different moments during the operating time the cells were allowed to recoverby turning off the bias voltage Recovery of the PL required several hoursMoreover as shown in Fig 113 only part of the PL could be recovered dependent
Fig 111 Time evolution of luminance of a stacked CP-based LEC at a constant current of167 mA cmminus2 The device was tested multiple times after storage for 17 days since the last run at167 mA cmminus2 Delay and heating applied to the cell are indicated in the graph The images of thecell to the right show (top) fluorescent image after run 7 before heating (middle) fluorescent imageafter run 7 and 7 h of heating at 60 degC (bottom) electroluminescent image after run 8 and 7 h ofheating at 60 degC (reproduced from Ref [100])
24 S van Reenen and M Kemerink
Fig 112 Design and materials for chemically fixed heterojunctions Top molecular structures of(left) PFP-BMes and (right) PFP-F Central pristine device with all ions in the FPF-F layer (leftdark blue) under an applied bias (central) the mobile fluoride anions (yellow) move into theFPF-BMes layer (light blue) where also electronic carriers are injected from the cathode (gray)and anode (white) the ions are compensated by injected charge carriers creating a p-n junction(right) A new immobile borate species (red) is formed (reproduced from Ref [95])
1 Light-Emitting Electrochemical Cells hellip 25
on the operating time This further proves that reversible and irreversible processesoccur simultaneously and quench the luminescence in these LECs
In the literature various irreversible processes have been observed and studied thatlead to reduced performance of LECsHerewewill shortlymention themain processes
Fig 113 Visualization of the reversible recovery and irreversible loss of PL intensity in a stackediTMC-LEC as a function of operating time The lines serve as a guide to the eye (reproduced fromRef [52])
Fig 114 Time evolution of luminance of a stacked CP-based LEC at a constant current of167 mA cmminus2 The operating time on the horizontal axis indicates accumulated run time underbias The cell was stored for 30 days in a N2-filled glovebox at room temperature after each runThe first run started right after deposition of the top Al electrode Photographs of theelectroluminescence of the LECs were taken at the end of the runs under which they are shown(reproduced from Ref [101])
26 S van Reenen and M Kemerink
Black spot formation is typically observed in LECs after prolonged operation asshown in Fig 114 [101] In the results shown here the irreversible decay ofluminance is concomitant with black spot formation in the polymer film The blackspots lead to an effective reduction of the emitting area In the same paper and in afollow-up work [102] the authors show that black spots are formed in unbiasedcells only if the cells are stored after deposition of the top Al electrode indicatingthat black spots do not necessarily appear during operation of LECs AlTal et al[102] found that black spots in EL also appeared in PL after prolonged operationHowever after switch-off of the bias voltage the black spots in the PL wereobserved to disappear Together these results strongly indicate that black spots maybe due to heavy doping promoted by chemical changes occurring at thecathodepolymer interface The exact origin of black spots formation in LECs hashowever not been identified Similar black spots have also been observed in LECscomprising iTMCs by Kalyuzhny et al [71] The researchers found that near thecathode where light emission takes place a quencher is formed by side reactionswith the iTMC that are assisted by moisture or oxygen
Another cause of irreversible degradation in LECs is a side-reaction with the iontransport material PEO In case of typical CP-based LECs based on PPV and PEOthe reduction level of the electrolyte lies below the conduction band of the CP (seeFig 115a) Consequently injection of electrons results in reduction of either theelectrolyte or the CP [76 103] Although the former is energetically favorable thelatter is kinetically preferred as electronic carrier transport though the electrolyte islow due to the electrically insulating character of PEO Nevertheless side reactionswith the PEO are likely to occur at the cathode interface as evidenced in the planarLEC shown in Fig 115b Such reactions may hamper electron injection causingan imbalance in carrier transport through the active layer Such imbalance canultimately lead to microshorts [104] when one of the doped regions grows com-pletely from one electrode towards the other Side reactions with PEO may also be
Fig 115 Left Schematic electron-energy level diagram for an LEC with the reduction level forthe PEO + KCF3SO3 electrolyte positioned within the bandgap of the CP (MEH-PPV) RightOptical microscopy images of the anodic (left) and cathodic (right) interfaces after 12 h ofoperation at V = 30 V and T = 360 K of a planar CP-based LEC with a 1 cm interelectrodespacing The line just left of the cathode appears to be due to cathodic electrochemical sidereactions (reproduced from Ref [103])
1 Light-Emitting Electrochemical Cells hellip 27
related to black spot formation as discussed above Lowering of the conductionband of the CP by use of different materials is a route to avoid such side reactionsand improve the overall performance of LECs [105] Use of alternativeion-solvating polymers has led to improved lifetime of LECs by improvement ofthe electrochemical stability window of the electrolyte [85ndash88]
14 Steady-State Phenomena
141 Potential and Ion Distribution
The potential profile in an LEC can be regarded as the fingerprint of the operatingmechanism that the LEC follows The ion distribution furthermore determines thepotential profile dependent on the formation of electric double layers and elec-trochemically doped regions Characterization of either of the two therefore givescrucial information on the operation of the device see also Table 12
1411 EDM
Stacked LECs typically have a too small interelectrode distance to enable surfacetechniques like electrostatic force microscopy (EFM) or scanning Kelvin probemicroscopy (SKPM) to directly measure the potential distribution in them Planarcells are however sufficiently large to allow such techniques to extract usefulinformation on the potential distribution During EFM a conductive tip is scannedin non-contact mode over an area without a feedback mechanism on the potentialThis is opposite to SKPM where a DC bias feedback is in place which keeps theelectrostatic potential difference between the conductive tip and the sample at zero
For LECs following the EDM the potential distribution is dominated by largepotential drops at the electrode interfaces and a field-free bulk [47 51] Figure 116shows the device layout and the typical potential distribution in an LEC followingthe EDM from EFM measurements on a planar Ru-based iTMC-LEC [51] InCP-based LECs similar potential profiles were obtained by SKPM in an N2-filledglovebox only after allowing the contacting Al electrodes to oxidize by contact toair [47] Consequently the appearance of the EDM model is assigned to deviceswith relatively poor carrier injection properties
The experiments show that the potential drop in LECs following the EDM ismainly located at the electrode interfaces due to formation of ionic space charge Asa result the bulk of the active layer becomes field-free Numerical modeling bydeMello et al [41ndash43] and van Reenen et al [47] confirm this picture and confirmthe requirement of relatively weak carrier injection to get into the EDM operatingregime Due to the absence of significant electrochemical doping the majority ofmobile ions remains distributed throughout the bulk of the active layer resulting inthe absence of net ionic charge
28 S van Reenen and M Kemerink
1412 ECDM
Matyba et al first reported on the potential distribution in planar CP-based LECsfollowing the ECDM as shown in Fig 117 [14] Figure 117b clearly demonstratesa significant potential drop in the bulk of the device that indicates a (dynamic) p-i-njunction Immediately after release of the bias voltage in this cell ie atopen-circuit conditions a built-in potential of 15 V can be observed at the sameposition (see Fig 117c) which is slightly smaller than the bandgap of the semi-conductor used in this cell This built-in voltage indicates the presence of a con-nection between an n-type doped semiconductor (on the left) and a p-type dopedsemiconductor (on the right)
In Fig 117b no potential drops are observed at the interfaces In otherexperiments however such potential drops have been observed upon use of Alinjecting contacts [47] This discrepancy is suggested to be due to the formation of athin layer of ion-containing material on top of the Au electrodes that screens someof the potential [14] The formation of this thin layer may be due to ions diffusingthrough the Au electrodes which is not possible in Al electrodes the potentialdrops observed at Al electrode interfaces were found to be similar to the expectedenergetic injection barriers determined from the difference in electrode Fermi-leveland the positions of the HOMO and LUMO levels of semiconductor
Au Au
Substrate
Fig 116 Top Schematic diagram of an unpatterned planar iTMC-LEC Bottom Experimentalcharacterization of the time dependence (green to red spaced by equal increments of 15 min) ofthe potential profiles for an Au[Ru(bpy)3][PF6]2Au device at Vbias = 5 V by EFM The electrodepositions are indicated by the dashed lines (reproduced from Ref [51])
1 Light-Emitting Electrochemical Cells hellip 29
Similar potential profiles relating to the ECDM have also been observed inplanar iTMC-LECs [53]
A technique that can be used to experimentally determine the ion distribution instacked LECs is time-of-flight secondary ion mass spectroscopy (ToF-SIMS) [106107] ToF-SIMS basically consists of sputtering the sample with a focused ionbeam while monitoring the secondary ions This way a depth profile of systemcomponents can be determined which can give information on the ion distributionin LECs One of the difficulties of this technique is that the ions used for sputteringinduce positive charges on the thin film surface [108] The induced charges giverise to large electric fields which disturb the ionic charge distribution in LECsstudied by ToF-SIMS Nevertheless side-by-side comparison between devices canstill give information on ionic charge distributions in LECs [107] ToF-SIMS wascarried out on stacked CP-based LECs as shown in Fig 118 [107] The results inFig 118b show first of all that an unbiased device (red linesymbols) suffers fromthe charge redistribution caused by the sputtering process Comparison with devicesbiased at 7 V for 2 min (blue linesymbols) and for 3 min (green linesymbols)however shows that because of prior device operation Li+ and CF3SO3
minusions have
Fig 117 Experimental SKPM data illustrating the potential distribution in a planar CP-basedLEC following the ECDM showing 2D topography image by atomic force microscopy (top left)electrostatic potential profile during steady-state operation at Vbias = +5 V (central left) transientpotential profile measured with the device disconnected (bottom left) directly after operation atVbias = +5 V micrograph showing light emission from the same device during steady-stateoperation at Vbias = +5 V(right) (reproduced from Ref [14])
30 S van Reenen and M Kemerink
moved towards the cathode and anode respectively This behavior is in line withthe ECDM
The combined results on potential and ion density profile distributions show thatLEC operating according to the ECDM separate anions from cations to formelectric double layers at the interfaces and doped regions in the bulk The dopedregions are separated by a low-conductive intrinsic region Consequently sharppotential drops are observed at the interfaces and in a small region in the bulkNumerical calculations confirm these experiments [38 47]
1413 PECDM
Potential profiles of LECs following the PECDM were first reported by Pingreeet al [45] and later on by Rodovsky et al [46] as shown in Fig 119 The majorityof the potential drops at the injecting contact that has the largest barrier for carrierinjection For the device shown in Fig 119 low-work function Ca electrodes wereused to contact the active layer to inject electrons and holes in the polymerMDMO-PPV Use of Ca leads to a large injection barrier for hole injection and a
Fig 118 Left Schematics illustrating a standard ToF-SIMS measurement on a LEC devicecomprising the salt Li+CF3SO3
minus Right ToF-SIMS profiles of Li+ (lines) and Fminus (circles Fminus
originates from CF3SO3minus) for dynamic junction LECs with a 100 nm thick active layer charged at
7 V for 3 min (blue) 2 min (green) and uncharged (red) (reproduced from Ref [107])
Fig 119 ExperimentalSKPM data illustrating thepotential distribution in aplanar CP-based LEC with Caelectrodes operated atVbias = 5 V following thePECDM The dashed linesindicate the electrodepositions (reproduced fromRef [46])
1 Light-Emitting Electrochemical Cells hellip 31
small injection barrier for electron injection As a result the active layer is pre-dominantly n-type doped and the light-emitting junction is positioned at the contactwith the largest injection barrier being the anode Therefore this device showspreferred n-type doping Devices with preferred p-type doping have also beenreported [45]
No numerical modeling has been reported in which potential profiles are cal-culated in devices following the PECDM
142 Position and Width of the Recombination Zone
1421 Studies in Planar LECs
Properties of the recombination zone depend strongly on the operation mechanismin LECs These have been studied extensively in planar LECs by photographs ofthe device during operation For planar cells operating in the EDM or the PECDMthe recombination zone is observed to sit close to one or even both of the electrodeinterfaces [46 51] For planar cells operating in the ECDM however the positionof the light-emitting junction can be found anywhere in the bulk of the devicedependent on properties related to carrier injection [45 46] and doping-dependenttransport [78 79] Moreover as these properties change during the whole elec-trochemical doping process of the active layer the junction position is also found tomove during operation [53] For device performance control of the junctionposition is important as to avoid the junction to lie close to the electrodes wherestrong EL quenching takes place
The initial position of the light-emitting junction is determined by the timerequired to form Ohmic contacts at both electrodes and the doping front propa-gation Ohmic contact formation takes longer in case a larger injection barrier needsto be overcome by EDL formation [45 46] Immediately when an Ohmic contact isformed carriers are injected which initiates the doping front propagation towardsthe opposite electrode Doping front propagation itself depends strongly on themobility of the electronic carriers that must be supplied to the front for electro-chemical doping [79] PPV-based planar CP-based LECs with Au electrodes have arelatively large injection barrier for electrons compared to holes as well as a holemobility that is one order of magnitude above the electron mobility [109] Hencethe initial position of the light-emitting junction in these devices is typically foundnear the cathode [46] Under such conditions distinguishing ECD and PECDMfrom ECDM behavior can be tricky The position of the junction region canhowever shift after initial formation because of continuous electrochemical doping[53] The final position of the junction region is then strongly determined by therespective electronic conductivity of the n-type and p-type doped regions whichmust lead to balanced transport of electrons and holes towards the recombinationzone [47 53 77 78]
32 S van Reenen and M Kemerink
The width of the junction in planar cells is found to range from 1ndash3 lmindependent of the interelectrode distance [13 46]
1422 Electrical Impedance Spectroscopy
Studies on stacked LECs often employ electrical impedance spectroscopy (EIS) inwhich the large differences in time scales associated with the various physical andchemical processes are used to separate these processes In particular EIS measuresthe dielectric properties of a medium as a function of frequency It is based on theperturbation of a quasi-equilibrium state by an AC bias voltage Vac This pertur-bation must be sufficiently small so that a linear response can be assumed Theperturbation results in a change in current that follows the oscillating bias voltage ~Vwith a phase difference Measurement of this oscillating current ~I allows determi-nation of the complex admittance Y
Y frac14~I~Vfrac14 Gthorn jC eth119THORN
with G the (real) conductance and C the (real) capacitance Dependent of the deviceunder test the overall conductance and capacitance are however the result of acomplex combination of various processes in the device Hence interpretation ofEIS is generally not straightforward and should therefore preferably be accompa-nied by modeling This modeling can be either by the use of equivalent circuits [84]or a numerical device model [63] based on the transport equations described inSect 125 In particular interpretation of EIS in LECs remains complicated due tothe presence of four types of charge carriers each with differentposition-dependent densities and mobilities Furthermore effects from binding andrecombination between these carriers as described in Sect 1253 will furthercomplicate matters
Studies reported on impedance spectroscopy in LECs have shown that thistechnique allows determination of ion conductivity and electric double layer for-mation in unbiased cells as well as determination of the recombination zone widthin biased cells that show light emission [63 84 93] It is anticipated that moreinformation can be obtained from biased LECs however this has yet to be con-clusively explored and reported
1423 Studies in Stacked LECs
In LECs the recombination zone is intrinsic and is therefore a relativelylow-conductive region sandwiched by relatively high-conductive doped regionsThe light-emitting p-i-n junction can thus in first-order approximation be con-sidered as a parallel plate capacitor The junction width is then
1 Light-Emitting Electrochemical Cells hellip 33
Lj frac14 e0er=C eth120THORN
where C is the areal capacitance in F mminus2 In EIS measurements the relatively slowions typically dominate at low frequencies below 10 kHz whereas electronicprocesses dominate above this frequency Therefore the capacitance for frequen-cies above 10 kHz is related to the junction hence the junction width can beestimated by this technique [54 63 84 110 111] Campbell et al modeledimpedance spectra extracted from stacked CP-based LECs and found that thejunction width depends on the bias voltage as shown in Fig 120 [111] At theturn-on of the cell above 17 V the junction width is observed to decreasewhereas the junction potential is observed to increase The width of the junction inthis stacked CP-based LEC decreases to values as low as 15 nm which is muchlower than the micron-sized widths found in planar CP-based LECs [13 46]
The low conductivity and the absence of doping in the junction region lead to theformation of space charge during operation Lenes et al [54] found that the current
Fig 120 Calculated potential drop across the junction (solid line) and junction width (dashedline) as a function of applied bias (reproduced from Ref [111])
Fig 121 I-V characteristics (left) as well as capacitance (filled symbols) and junction thickness(empty symbols) measurements (right) at different times during Vbias = 35 V operation of astacked Ir-based iTMC-LEC The black lines in a indicate a quadratic dependence (reproducedfrom Ref [54])
34 S van Reenen and M Kemerink
through this region is space charge limited as shown by the quadratic dependence ofthe current in Fig 121a Space charged limited current can be described by
JSCLC frac14 aV2
L3j eth121THORN
where a is a coefficient that depends on the dielectric constant carrier mobilitiesand bimolecular recombination rate [112] The increase in current during fixedvoltage operation in Fig 121a can therefore be explained by a continuous reduc-tion in junction width Lj This is confirmed by determination of the junction widthin similar cells under similar conditions by characterization of the high frequencycapacitance as shown in Fig 121b This continuous reduction of the junction widthis due to continuous doping of the active layer
143 Current-Voltage Characteristic
Measurement of a general I-V characteristic in LECs is complicated because of thetypical hysteresis due to on-going reversible and irreversible reactions [100 101]combined with the large variation in turn-on times [55] Hence a difference involtage sweep rate can have dramatic impact on the resultant I-V characteristic Thisprompts the need to always report time-dependent characteristics of currentluminance and efficiency next to voltage-dependent characteristics [63 64]Nevertheless experimental data in the literature still show resemblance between I-V characteristics for various materials used in LECs A few examples of experi-mentally determined I-V characteristics are shown in Fig 122
Understanding of the I-V characteristics in LECs necessitate numerical modelingdue to the large amount of processes happening simultaneously in LECs Suchnumerical modeling efforts were initiated by D L Smith who used a steady-statemodel to describe carrier transport in LECs [38] The model predicted that the
Fig 122 I-V characteristics of stacked LECs based on a multifluorophoric conjugated copolymermixed with a LiCF3SO3 + trimethylolpropane ethoxylate electrolyte (left reproduced from Ref[113]) a MDMO-PPV CP mixed with a KCF3SO3 + poly(ethylene oxide) electrolyte (centralreproduced from Ref [63]) and a Tris(22rsquo-bipyridine)ruthenium(II) complex (straight linestested in drybox dashed lines tested in air) (right reproduced from Ref [71])
1 Light-Emitting Electrochemical Cells hellip 35
electrochemical doping density depends exponentially on the applied bias voltagefor voltages near the band gap of the semiconductor As a result the current growsexponentially for increasing bias voltage as shown in Fig 123a Such exponentialgrowth is also observed in experiments as shown in Fig 122b and c for biasvoltages around the bandgap of the semiconductor ie roughly 2 eV
The experiments displayed in Fig 122b show that at relatively high bias volt-ages the current levels off Modeling by Manzanares et al showed that the I-V curve levels off at high doping densities as shown in Fig 123b [39] This is dueto all ions being used for electrochemical doping preventing further doping at evenhigher bias voltages According to their model any additional voltage is added tothe interfaces Other work including modeling and experiments [47] have shownthis not to be the case the additionally applied bias voltage drops in the junctionregion The enhanced field in the junction can furthermore lead to a broadening ofthe recombination zone due to the requirement for current conservation [63]
Mills and Lonergan developed a numerical and analytical model that describesthe complete I-V characteristic across the whole possible bias voltage range asshown in Fig 123c [48] They discern four regimes In the first low injectionregime the current grows exponentially due to increased carrier injection by EDLformation Here the bulk is field-free and the operation of the LEC follows theEDM Increasing the bias voltage gets the LEC in the second space charge regimewhere one type of electronic carrier is injected efficiently In this regime the devicefollows the PECDM The space charge of this majority carrier in the device thenaffects the device potential in such a way that further injection is retarded leading toonly a weak increase of the current density Simultaneously increase of the biasvoltage within this regime leads to enhancement of the injection of the minoritycarriers Increasing the bias even further gets the cell in the third bipolar injectionregime where carrier injection for both electron and hole injection is efficientallowing anions and cations to separate to form electrochemically doped regionswhile the current again increases exponentially with bias At these bias voltages thedevice follows the ECDM Further enhancement of the bias voltage will ultimatelylead to complete exhaustion of the available mobile ions so the device enters thefourth high injection regime At this point enhancement of the bias voltage does
Fig 123 Calculated I-V characteristics in LECs by Smith (Left reproduced from Ref [38])Manzanares et al (central reproduced from Ref [39]) and Mills and Lonergan (right reproducedfrom Ref [48])
36 S van Reenen and M Kemerink
not enhance the doping density in the device anymore resulting in a much weakerincrease in current density Any additional voltage is then dropped predominantlyover the junction region and to a minor extent at the interfaces to enhance carrierinjection to accommodate the increase in current density
The bias voltage at which each regime occurs depends strongly on the availablemobile ion densities and the injection barriers for electrons and holes Hence largevariations are present between I-V characteristics in different LECs
144 Luminescence Quenching and Reabsorption
Mobile ions in LECs enable efficient carrier injection and transport by EDL for-mation and electrochemical doping at relatively low bias voltages However toachieve an efficient electroluminescent device the resultant large density of currentmust be converted into excitons that subsequently decay radiatively generatingphotons that leave the device without being reabsorbed
To achieve a maximal recombination efficiency of electrons and holes intoexcitons it is required that injected electrons and holes do not leave the active layerat the opposite electrode ie the anode and the cathode respectively This is thecase if LECs operate in the ECDM and both n- and p-type electrochemical dopinghas developed in such a way that the recombination zone is sufficiently far awayfrom the electrodes High doping densities should then make it impossible forelectrons or holes to move through oppositely doped regions without recombining[47]
The fraction of excitons that decays radiatively depends first of all on the type ofemitter that is used and the distribution in exciton spin-states Conjugated polymerslike PPV are singlet emitters whereas eg Ir-based iTMCs are triplet emittersSecond it depends on the rates of non-radiative decay processes that compete withradiative decay Excitons can eg be quenched by the presence of large electricfields [114] other excitons [115] or polarons [116] Exciton quenching by polaronshas been found to contribute significantly to loss in efficiency in LECs [64 117]The reason is the relatively large density of electrochemical doping next to arelatively thin recombination zone eg 15 or 22 nm thick as shown inFigs 120 and 121b Exciton quenching by polarons occurs through diffusion ofexcitons through the semiconductor during their lifetime followed by either of twopossible processes Foumlrster resonance energy transfer (FRET) or charge transfer(CT) [117] Both processes are depicted in Fig 124 In case of FRET the excitontransfers its energy to a polaron leading to the loss of the exciton and an excitedpolaron that will relax and generate heat In case of CT the polaron recombineswith the exciton leading to the loss of the exciton and the generation of a newpolaron
The effect of these quenching mechanisms becomes stronger for higher dopingdensities In previous paragraphs we saw that doping densities increase in time(Sect 131) and for enhanced voltages (Sect 143) Indeed in operational LECs
1 Light-Emitting Electrochemical Cells hellip 37
the efficiency is found to typically roll-off in time (see also Fig 15 in Sect 131)and for enhanced voltages [55] Integrating the above-mentioned quenchingmechanisms into a device model allows to reproduce this roll-off see alsoSect 146 [64]
Following radiative decay the generated photons need to leave the deviceHowever Kaihovirta et al found that a significant fraction of the generated photonscan be reabsorbed by electrochemical doping in CP- and iTMC-based LECs [118119] Electrochemical doping namely leads to an additional absorption band atsimilar wavelengths as the emitting transition shown for example in Fig 125 Fora CP-based LEC based on Super Yellow PPV Kaihovirta found that 100 nm ofactive layer leads to 10 reabsorption whereas 10 lm of active layer leads toover 70 reabsorption The effect depends on the overlap between the emissionspectrum and the self-absorption spectrum induced by doping Hence this effect canvary strongly for different materials The reabsorption in a yellow-emitting Ir-basediTMC-LEC was found to be significantly less 4 in a 95 nm thick device and 40
Fig 124 Schematic representation of exciton quenching mechanisms in LECs Left excitondiffusion follow by Foumlrster resonance energy transfer Right exciton diffusion followed by chargetransfer (reproduced from Ref [117])
Fig 125 Absorption and ELintensity in a stackedCP-based LEC Theabsorption spectra are shownfor a doped and undopeddevice as indicated(reproduced from Ref [118])
38 S van Reenen and M Kemerink
in a 10 lm thick device This shows that thin devices are likely preferred toachieve efficient LECs It also suggests a materials optimization target
145 Color Tuning and Cavity Effects
Color tuning in electroluminescent devices based on organics is relatively easycompared to those based on inorganics due to the large number of materials withdifferent bandgaps available through synthesis In polymers and small moleculesthe bandgap can be tuned by using different repeat units or by adding differentelectron donatingaccepting moieties in the periphery of the compound respectively[113 120] In iTMCs variation of the organic ligands allows coverage of thecomplete visible light emission spectrum despite the limited possibilities in vari-ation of the metal complex [121] Finally the color of nanoparticles and quantumdots are mainly related to their size as a consequence of the quantum confinementfacilitating color tuning in LECs based on these emitters These aspects are furtherdescribed in the section III of the book
Besides color tuning through materials design the emission spectrum of LECscan also be modified through photonic effects since the thickness of the active layerin stacked LECs is comparable to the wavelengths of the emitted photonsMoreover typically one reflective electrode is used to contact the device Thiscombination gives stacked LECs a microcavity structure that through interferenceeffects can have a significant influence on the emission spectrum and must therefore
Fig 126 Simulated and measured EL spectra of a stacked 450 nm thick iTMC-LEC at 8 (topleft) 12 (top right) 18 (bottom left) and 58 (bottom right) min after a bias of 25 V was appliedThe recombination zone position (zi) with respect to the cathode position was determined by fittingthe simulated and measured EL spectra and is shown in each figure (reproduced from Ref [127])
1 Light-Emitting Electrochemical Cells hellip 39
be taken into account when optimizing LECs These cavity effects are known to beimportant in OLEDs [122ndash124] and in organic photovoltaic cells [125]
Although Chap 3 is devoted to color tuning by cavity effects we will brieflycomment on it here The position where recombination takes place in electrolu-minescent devices has a strong effect on the emission spectrum [124] Wang and Sutook advantage of cavity effects in 450 nm thick LECs to study the migration of thelight-emitting junction in time as shown in Fig 126 [126] They found that thejunction first formed closest to the electrode with the largest injection barrier whichis the cathode in this case (see Fig 126a) This is due to more ions being requiredto turn the cathode into an Ohmic contact In time however this Ohmic contact isformed resulting in balanced carrier injection that leads to a junction shift towardsthe center of the device (see Fig 126d) These results show that effects of carrierinjection and transport on the junction position during operation can be studiedusing microcavity effects in stacked LECs We are not aware of non-diagnosticapplication of these effects in color tuning of LECs
146 Efficiency Values and Limits
Comparison of performance between LEC devices is extremely complicatedEfficiencies as high as 398 lmW [128] luminances above 10000 cdm2 [129] andlifetimes of 3000 h [130] (defined here as the time for the brightness to decay tohalf-maximum) have been reported for LECs However these values do not applyto a single device and are all measured under different experimental conditions Forexample the efficiency of LECs depends strongly on the emission wavelength aswell as the brightness and time at which the efficiency was measured [131]
The time dependence of operation of LECs has already been discussed in pre-vious paragraphs As a virtually unavoidable result of the device physics variationsin efficiency over time are expected due to eg continuous electrochemical dopingand changes in the recombination zone width and position It has however beenshown by Tordera et al that the characteristic efficiency roll-off of LECs (seeSect 13) can be significantly suppressed by a pulsed instead of a DC drivingscheme [94]
The tradeoff between efficiency and brightness has been discussed quantitativelyin Ref [64] Enhancement of the brightness in an LEC is achieved by improving thecurrent density which in turn is achieved by enhancing the doping densityHowever this enhanced doping density simultaneously boosts the rate ofexciton-polaron quenching by Foumlrster resonant energy transfer (FRET) andorcharge transfer (CT) as discussed in Sect 144 This quenching results in a low-ering of the electroluminescent efficiency [64] To avoid this tradeoff it is requiredto (1) reduce FRET by reduction of the overlap between exciton emission spectraand polaron absorption spectra and (2) reduce CT by suppression of exciton dif-fusion by eg introduction of structural disorder Unfortunately the latter would
40 S van Reenen and M Kemerink
lead to (even) lower charge carrier mobilities For iTMC complexes the structuraldisorder can possibly be tuned through the addition of large side-chains [132]
15 Conclusion and Outlook
Work carried out by various groups around the world has led to a coherentunderstanding of the complicated device physics of LECs The operation of LECsstrongly depends on carrier injection in the active layer which is predominantlydetermined by the injection barrier and the applied bias voltage The preferredoperational mechanism of LECs is the electrochemical doping model that occurs incase carrier injection does not limit device operation leading to electrochemicaldoping of the active layer The process of electrochemical doping in LECs governsthe transient properties relating to turn-on hysteresis and lifetime as well as thesteady-state properties relating to potential and carrier distribution carrier recom-bination and electroluminescence quenching
As the majority of the device physics is now relatively well understood it seemslogical to direct further efforts towards materials development to limit electrolu-minescence quenching self-absorption and degradation which lead to significantlosses in performance Electroluminescence quenching can be reduced by use ofdyes that limit exciton diffusion towards polarons or by enhancement of theradiative decay rate Self-absorption can be reduced by use of relatively thin activelayer films and improved control of the position of the recombination zone Thelatter may also be used to tune the emission spectrum and minimize its overlap withquenching or absorbing transitions see Sects 144 and 145 Regarding degra-dation its origins are yet to be fully understood Nevertheless it seems likely thation-solvating polymers with a larger electrochemical stability window are desired
To accommodate materials research and device development the field wouldgreatly benefit from standardized tests on reporting performance characteristics ofLECs The extensively discussed transient behavior of LECs makes their perfor-mance strongly dependent on measurement time and driving conditions oftenmaking it nearly impossible to compare reported performance indicators As a firststarting point peak light outputs and efficiencies should always be accompanied byinformation about their transients such as the turn-on time and the time to half-peakvalue Setup of a standardized characterization protocol is however far fromstraightforward because of the multidimensional parameter space of light-emittingelectrochemical cells
References
1 A Bernanose M Comte P Vouaux J Chim Phys 50 64 (1953)2 CW Tang SA VanSlyke Appl Phys Lett 51 913 (1987)
1 Light-Emitting Electrochemical Cells hellip 41
3 JH Burroughes DDC Bradley AR Brown RN Marks K Mackay RH Friend PLBurns AB Holmes Nature 347 539 (1990)
4 DR Baigent NC Greenham J Gruumlner RN Marks RH Friend SC Moratti ABHolmes Synth Met 67 3 (1994)
5 K Fesser AR Bishop DK Campbell Phys Rev B 27 4804 (1983)6 A Koumlhler H Baumlssler Electronic Processes in Organic Semiconductors An Introduction
(Wiley 2015)7 PWM Blom MCJM Vissenberg Mater Sci Eng R Rep 27 53 (2000)8 H Imahori H Norieda Y Nishimura I Yamazaki K Higuchi N Kato T Motohiro H
Yamada K Tamaki M Arimura Y Sakata J Phys Chem B 104 1253 (2000)9 P Andrew Science 290 785 (2000)
10 E Dulkeith AC Morteani T Niedereichholz TA Klar J Feldmann SA Levi FCJMVan Veggel DN Reinhoudt M Moumlller DI Gittins Phys Rev Lett 89 (2002)
11 S Reineke F Lindner G Schwartz N Seidler K Walzer B Luumlssem K Leo Nature 459234 (2009)
12 M Mesta M Carvelli RJ de Vries H van Eersel JJM van der Holst M Schober MFurno B Luumlssem K Leo P Loebl R Coehoorn PA Bobbert Nat Mater 12 652 (2013)
13 Q Pei G Yu C Zhang Y Yang AJ Heeger Science 269 1086 (1995)14 J-K Lee DS Yoo ES Handy MF Rubner Appl Phys Lett 69 1686 (1996)15 ES Handy AJ Pal MF Rubner J Am Chem Soc 121 3525 (1999)16 FG Gao AJ Bard J Am Chem Soc 122 7426 (2000)17 H Rudmann MF Rubner J Appl Phys 90 4338 (2001)18 AJ Norell Bader AA Ilkevich IV Kosilkin JM Leger Nano Lett 11 461 (2011)19 S Tang W-Y Tan X-H Zhu L Edman Chem Commun 49 4926 (2013)20 A Pertegaacutes D Tordera JJ Serrano-Peacuterez E Ortiacute HJ Bolink J Am Chem Soc 135
18008 (2013)21 C Bizzarri C Strabler J Prock B Trettenbrein M Ruggenthaler C-H Yang F Polo A
Iordache P Bruumlggeller LD Cola Inorg Chem 53 10944 (2014)22 G Qian Y Lin G Wantz AR Davis KR Carter JJ Watkins Adv Funct Mater 24
4484 (2014)23 D Asil JA Foster A Patra X de Hatten J del Barrio OA Scherman JR Nitschke R
H Friend Angew Chem Int Ed 53 8388 (2014)24 S Keller EC Constable CE Housecroft M Neuburger A Prescimone G Longo A
Pertegaacutes M Sessolo HJ Bolink Dalton Trans 43 16593 (2014)25 MY Wong GJ Hedley G Xie LS Koumllln IDW Samuel A Pertegaacutes HJ Bolink E
Zysman-Colman Chem Mater 27 6535 (2015)26 MF Ayguumller MD Weber BMD Puscher DD Medina P Docampo RD Costa
J Phys Chem C 119 12047 (2015)27 H Zhang H Lin C Liang H Liu J Liang Y Zhao W Zhang M Sun W Xiao H Li S
Polizzi D Li F Zhang Z He WCH Choy Adv Funct Mater 25 7226 (2015)28 MS Subeesh K Shanmugasundaram CD Sunesh TP Nguyen Y Choe J Phys Chem
C 119 23676 (2015)29 MD Weber V Nikolaou JE Wittmann A Nikolaou PA Angaridis G Charalambidis
C Stangel A Kahnt AG Coutsolelos RD Costa Chem Commun 52 1602 (2016)30 MD Weber M Adam RR Tykwinski RD Costa Adv Funct Mater 25 5066 (2015)31 S Jenatsch L Wang M Bulloni AC Veacuteron B Ruhstaller S Altazin F Nuumlesch R
Hany ACS Appl Mater Interfaces 8 6554 (2016)32 Z Zhang K Guo Y Li X Li G Guan H Li Y Luo F Zhao Q Zhang B Wei Q Pei
H Peng Nat Photonics (2015)33 A Asadpoordarvish A Sandstroumlm C Larsen R Bollstroumlm M Toivakka R Oumlsterbacka
L Edman Adv Funct Mater 25 3238 (2015)34 A Sandstroumlm A Asadpoordarvish J Enevold L Edman Adv Mater 26 4975 (2014)35 A Sandstroumlm L Edman Energy Technol 3 329 (2015)36 Q Pei Yang G Yu C Zhang AJ Heeger J Am Chem Soc 118 3922 (1996)
42 S van Reenen and M Kemerink
37 DL Smith J Appl Phys 81 2869 (1997)38 JA Manzanares H Reiss AJ Heeger J Phys Chem B 102 4327 (1998)39 JC deMello Phys Rev B 66 (2002)40 JC deMello N Tessler SC Graham RH Friend Phys Rev B 57 12951 (1998)41 JC deMello JJM Halls SC Graham N Tessler RH Friend Phys Rev Lett 85 421
(2000)42 JM Leger SA Carter B Ruhstaller J Appl Phys 98 124907 (2005)43 LSC Pingree DB Rodovsky DC Coffey GP Bartholomew DS Ginger J Am
Chem Soc 129 15903 (2007)44 DB Rodovsky OG Reid LSC Pingree DS Ginger ACS Nano 4 2673 (2010)45 S van Reenen P Matyba A Dzwilewski RAJ Janssen L Edman M Kemerink J Am
Chem Soc 132 13776 (2010)46 TJ Mills MC Lonergan Phys Rev B 85 (2012)47 Q Pei AJ Heeger Nat Mater 7 167 (2008)48 J Gao J Dane Appl Phys Lett 84 2778 (2004)49 JD Slinker JA DeFranco MJ Jaquith WR Silveira Y-W Zhong JM
Moran-Mirabal HG Craighead HD Abruntildea JA Marohn GG Malliaras Nat Mater6 894 (2007)
50 P Matyba K Maturova M Kemerink ND Robinson L Edman Nat Mater 8 672(2009)
51 SB Meier D Hartmann D Tordera HJ Bolink A Winnacker W Sarfert Phys ChemChem Phys 14 10886 (2012)
52 SB Meier S van Reenen B Lefevre D Hartmann HJ Bolink A Winnacker W SarfertM Kemerink Adv Funct Mater 23 3531 (2013)
53 M Lenes G Garcia-Belmonte D Tordera A Pertegaacutes J Bisquert HJ Bolink AdvFunct Mater 21 1581 (2011)
54 S van Reenen T Akatsuka D Tordera M Kemerink HJ Bolink J Am Chem Soc 135886 (2013)
55 P Langevin in Recomb Mobilites Ions Dans Gaz Ann Chim Phys (1903) p 43356 W Shockley WT Read Phys Rev 87 835 (1952)57 RN Hall Phys Rev 87 387 (1952)58 M Kuik LJA Koster GAH Wetzelaer PWM Blom Phys Rev Lett 107 (2011)59 PR Emtage JJ OrsquoDwyer Phys Rev Lett 16 356 (1966)60 J C deMello J Comput Phys 181 564 (2002)61 D Tordera S Meier M Lenes RD Costa E Ortiacute W Sarfert HJ Bolink Adv Mater
24 897 (2012)62 S van Reenen RAJ Janssen M Kemerink Adv Funct Mater 22 4547 (2012)63 S van Reenen RAJ Janssen M Kemerink Adv Funct Mater 25 3066 (2015)64 J-H Shin ND Robinson S Xiao L Edman Adv Funct Mater 17 1807 (2007)65 J Fang Y Yang L Edman Appl Phys Lett 93 63503 (2008)66 Y Hu J Gao J Am Chem Soc 133 2227 (2011)67 Y Zhang J Gao J Appl Phys 100 84501 (2006)68 CH Lyons ED Abbas J-K Lee MF Rubner J Am Chem Soc 120 12100 (1998)69 HJ Bolink L Cappelli E Coronado M Graumltzel E Ortiacute RD Costa PM Viruela MK
Nazeeruddin J Am Chem Soc 128 14786 (2006)70 G Kalyuzhny M Buda J McNeill P Barbara AJ Bard J Am Chem Soc 125 6272
(2003)71 J Slinker D Bernards PL Houston HD Abruntildea S Bernhard GG Malliaras Chem
Commun 2392 (2003)72 AL Holt JM Leger SA Carter J Chem Phys 123 44704 (2005)73 J Gao J Dane J Appl Phys 98 63513 (2005)74 J-H Shin ND Robinson S Xiao L Edman Adv Funct Mater 17 1807 (2007)75 T Waringgberg PR Hania ND Robinson J-H Shin P Matyba L Edman Adv Mater 20
1744 (2008)
1 Light-Emitting Electrochemical Cells hellip 43
76 S van Reenen P Matyba A Dzwilewski RAJ Janssen L Edman M Kemerink AdvFunct Mater 21 1795 (2011)
77 S van Reenen RAJ Janssen M Kemerink Org Electron 12 1746 (2011)78 H Shimotani G Diguet Y Iwasa Appl Phys Lett 86 22104 (2005)79 VI Arkhipov EV Emelianova P Heremans H Baumlssler Phys Rev B 72 235202 (2005)80 ND Robinson J-H Shin M Berggren L Edman Phys Rev B 74 (2006)81 V Bychkov P Matyba V Akkerman M Modestov D Valiev G Brodin CK Law
M Marklund L Edman Phys Rev Lett 107 (2011)82 V Bychkov O Jukimenko M Modestov M Marklund Phys Rev B 85 (2012)83 A Munar A Sandstroumlm S Tang L Edman Adv Funct Mater 22 1511 (2012)84 S Tang L Edman J Phys Chem Lett 1 2727 (2010)85 S Tang J Mindemark CMG Araujo D Brandell L Edman Chem Mater 26 5083
(2014)86 J Mindemark S Tang J Wang N Kaihovirta D Brandell L Edman Chem Mater
(2016)87 J Mindemark L Edman J Mater Chem C 4 420 (2016)88 Y Shen DD Kuddes CA Naquin TW Hesterberg C Kusmierz BJ Holliday
JD Slinker Appl Phys Lett 102 203305 (2013)89 KJ Suhr LD Bastatas Y Shen LA Mitchell BJ Holliday JD Slinker ACS Appl
Mater Int 8 8888 (2016)90 RD Costa A Pertegaacutes E Ortiacute HJ Bolink Chem Mater 22 1288 (2010)91 J Fang P Matyba L Edman Adv Funct Mater 19 2671 (2009)92 G Yu Y Cao C Zhang Y Li J Gao AJ Heeger Appl Phys Lett 73 111 (1998)93 D Tordera S Meier M Lenes RD Costa E Ortiacute W Sarfert HJ Bolink Adv Mater
24 897 (2012)94 CV Hoven H Wang M Elbing L Garner D Winkelhaus GC Bazan Nat Mater
(2010)95 IV Kosilkin MS Martens MP Murphy JM Leger Chem Mater 22 4838 (2010)96 J Gao G Yu AJ Heeger Appl Phys Lett 71 1293 (1997)97 L Edman J Appl Phys 95 4357 (2004)98 Y Shao GC Bazan AJ Heeger Adv Mater 19 365 (2007)99 X Li J Gao G Liu Appl Phys Lett 102 223303 (2013)100 X Li F AlTal G Liu J Gao Appl Phys Lett 103 243304 (2013)101 F AlTal J Gao Org Electron 18 1 (2015)102 J Fang P Matyba ND Robinson L Edman J Am Chem Soc 130 4562 (2008)103 J-H Shin S Xiao L Edman Adv Funct Mater 16 949 (2006)104 P Matyba MR Andersson L Edman Org Electron 9 699 (2008)105 SB Toshner Z Zhu IV Kosilkin JM Leger ACS Appl Mater Interfaces 4 1149
(2012)106 TD Shoji Z Zhu JM Leger ACS Appl Mater Interfaces 5 11509 (2013)107 S Krivec T Detzel M Buchmayr H Hutter Appl Surf Sci 257 25 (2010)108 L Bozano SA Carter JC Scott GG Malliaras PJ Brock Appl Phys Lett 74 1132
(1999)109 Y Li J Gao G Yu Y Cao AJ Heeger Chem Phys Lett 287 83 (1998)110 IH Campbell DL Smith CJ Neef JP Ferraris Appl Phys Lett 72 2565 (1998)111 MA Lampert P Mark Current Injection in Solids (Academic Press 1970)112 J Kalinowski W Stampor J Szmytkowski D Virgili M Cocchi V Fattori C Sabatini
Phys Rev B 74 (2006)113 SM King D Dai C Rothe AP Monkman Phys Rev B 76 (2007)114 JM Hodgkiss S Albert-Seifried A Rao AJ Barker AR Campbell RA Marsh RH
Friend Adv Funct Mater 22 1567 (2012)115 S van Reenen MV Vitorino SCJ Meskers RAJ Janssen M Kemerink Phys Rev
B 89 (2014)116 N Kaihovirta A Asadpoordarvish A Sandstroumlm L Edman ACS Photonics 1 182 (2014)
44 S van Reenen and M Kemerink
117 N Kaihovirta G Longo L Gil-Escrig HJ Bolink L Edman Appl Phys Lett 106103502 (2015)
118 S Tang J Pan HA Buchholz L Edman J Am Chem Soc 135 3647 (2013)119 S Tang J Pan H Buchholz L Edman ACS Appl Mater Interfaces 3 3384 (2011)120 MS Lowry S Bernhard Chem - Eur J 12 7970 (2006)121 M-H Lu JC Sturm J Appl Phys 91 595 (2002)122 JM Leger SA Carter B Ruhstaller H-G Nothofer U Scherf H Tillman H-H
Houmlrhold Phys Rev B 68 (2003)123 SLM van Mensfoort M Carvelli M Megens D Wehenkel M Bartyzel H Greiner
RAJ Janssen R Coehoorn Nat Photonics 4 329 (2010)124 J Gilot I Barbu MM Wienk RAJ Janssen Appl Phys Lett 91 113520 (2007)125 T-W Wang H-C Su Org Electron 14 2269 (2013)126 HJ Bolink E Coronado RD Costa N Lardieacutes E Ortiacute Inorg Chem 47 9149 (2008)127 Y Xiong L Li J Liang H Gao S Chou Q Pei Mater Horiz 2 338 (2015)128 HJ Bolink E Coronado RD Costa E Ortiacute M Sessolo S Graber K Doyle
M Neuburger CE Housecroft EC Constable Adv Mater 20 3910 (2008)129 T Hu L He L Duan Y Qiu J Mater Chem 22 4206 (2012)130 RD Costa E Ortiacute HJ Bolink S Graber CE Housecroft EC Constable Adv Funct
Mater 20 1511 (2010)131 JC deMello Nat Mater 6 796 (2007)132 H-C Su J-H Hsu Dalton Trans 44 8330 (2015)
1 Light-Emitting Electrochemical Cells hellip 45
Part IIDefinition and Role of the Ionic
Additives
Chapter 2Optical-Beam-Induced-Current Imagingof Planar Polymer Light-EmittingElectrochemical Cells
Faleh AlTal and Jun Gao
Abstract In this chapter we describe optical-beam-induced-current (OBIC) andscanning photoluminescence (PL) imaging of extremely large planar LECs thathave been frozen to preserve the doping profile This complements the basicsdescribed in Chap 1 with respect to device mechanism and characterization Wesucceeded in resolving the depletion width for the first time of a frozen LEC p-n junction and a frozen LEC p-i-n junction These optical scanning results reveal asurprisingly strong built-in potential that is independent of the electrode workfunction and an extremely narrow junction depletion region that is about 02 ofthe interelectrode spacing These findings provide new insight into the electronicstructure of the LEC junction Since only about 02 of the entire device area isphotoactive in response to an incident optical beam the effective junction width (orvolume) of polymer-based LECs must be dramatically increased to realize a moreefficient device
Keywords P-i-n junction P-n junction Electrochemical doping Optical-beam-induced current Light-emitting electrochemical cell
21 Polymer Light-Emitting Electrochemical Cells
211 Background
A light-emitting electrochemical cell (LEC) is a solid-state two-terminal device thatemploys a mixed ionicelectronic conductor as the active layermdashsee Chap 1 formore details [1ndash7] The first LECs were demonstrated by Pei et al and consisted ofa luminescent conjugated polymer as the emitter and a polymer electrolyte as theion conductor sandwiched between two electrodes [8] A direct current (DC) bias
F AlTal J Gao (amp)Department of Physics Engineering Physics and Astronomy Queenrsquos University 64 BaderLane Kingston ON K7L 3N6 Canadae-mail jungaoqueensuca
copy Springer International Publishing AG 2017RD Costa (ed) Light-Emitting Electrochemical CellsDOI 101007978-3-319-58613-7_2
49
comparable to the luminescent polymer energy gap is required to activate apolymer LEC Charge injection at the positive electrode (anode) and the negativeelectrode (cathode) leads to the oxidation and reduction of the luminescent polymerMeanwhile the mobile ions from the polymer electrolyte redistribute and com-pensate the injected electronic charges causing in situ electrochemical doping ofthe luminescent polymer The doped polymer is electrically neutral but has elevatedelectronic conductivity due to the extra injected electronic charges With remark-able insight it was hypothesized by the inventors of LECs that the in situ dopingwas a dynamic process doping initially occurred at the electrode-polymer inter-faces but the doped regions would expand until they meet to form a p-n junctionThe formation of the p-n junction opens up a continuous pathway for the electroniccharges namely electrons and holes which recombine in the junction region to giveoff light emission [9] The LEC operation mechanism as described above isdepicted schematically in Fig 21 Light emission in an LEC is the result ofradiative recombination of the injected electrons and holes in the vicinity of thejunction In this regard the polymer LEC is analogous to a conventional p-n junction light-emitting diode (LED) in that both contain a semiconductorhomojunction A polymer LEC is therefore fundamentally different from a polymerLED made with the same luminescent polymer [10]
The polymer-based LEC was developed to address two main drawbacks ofconventional polymer LEDs The active layer of a prototypical polymer LED is apristine light-emitting polymer film For visible light-emitting applications theenergy gap of the polymer needs to be between 16 and 31 eV Due to the largeenergy gap undoped polymer films have high resistivity and need to be very thin inorder to inject a sufficient amount of current The ultrathin (ca 100 nm) polymerfilm however is prone to pinholes and the effect of exciton quenching [11]Moreover the injection of electrons and holes needs to be balanced for maximumelectroluminescence (EL) efficiency To facilitate electrons injection a low workfunction reactive metal cathode is commonly used The highly reactive cathodematerial (such as calcium) increases the chance of device failure unless the polymerLED is carefully encapsulated The polymer LEC by contrast operates on in situelectrochemical doping of the luminescent polymer due to the presence of mobileions in the composite material The doped polymer is much more conductive than apristine one so that a thicker active layer could be used The high conductivity ofthe doped LEC film also ensures strong and balanced charge injection at theelectrode interfaces Regardless of the electrode work function efficient charge
Fig 21 Illustration of doping propagation and junction formation process in polymer-basedLECs h stands for holes e stands for electrons A stands for anions and C stands for cations
50 F AlTal and J Gao
injection can occur via quantum mechanical tunnelling between a metal electrodeand a heavily doped semiconductor Heavy doping is in fact how an ohmic contactis made in conventional inorganic semiconductor devices Indeed highly efficientsandwich LECs had been demonstrated very early on with air-stable cathodematerials and without express optimization of the active layer thickness [12ndash14]
The unique advantages of LECs have led to a renaissance of interest in thesedevices in recent years [2 15ndash23] LECs can be made with not only luminescentpolymers and polymer electrolytes but also ionic metal complexes or low-cost smallmolecules [24ndash32] The LEC operation mechanism is inherently complex due to thepresence of both ionic and electronic charges in a mixed ionicelectronic conductorThe fundamental operating mechanism of polymer-based LECs for example hadbeen a subject of intense debate [33 34] The aim of many studies both theoretical andexperimental had been to elucidate the basic processes of LECs and the properties ofthe LEC junctions [35ndash40] For polymer-based LECs it is established that in situelectrochemical doping and junction formation are the fundamental processes thatdictate the dynamic activationturn-on behaviour of LECs as well as the static junctionproperties However we still lack knowledge about the basic properties of an LEC p-n or p-i-n junction In this chapter we describe our recent experimental work on thescanning optical imaging of frozen planar LECs to probe the electronic structures of anLEC junction This is meant to complement the basics described in Chap 1 withrespect to device mechanism and characterization In brief we performed four con-secutive optical-beam-induced-current (OBIC) and scanning photoluminescence(PL) imaging of LECs with a planar (versus sandwich) configuration We devised fourdifferent scanning setups each with an increasing scanning resolution than the pre-vious one [41ndash44] We succeeded in resolving the depletion width for the first time ofa frozen polymer p-n junction and a polymer p-i-n junction These optical scanningresults reveal a surprisingly strong built-in potential that is independent of the elec-trode work functions and an extremely narrow junction depletion region that is lessthan 02 of the interelectrode spacing These findings have profound implications onthe development of more practical and efficient LECs as well as the fundamentalscience of mixed conductorsmdasheg see Chaps 4 5 6 and 10 The scanning mea-surements were performed on extremely large frozen-junction planar LECs Gao andhis colleagues were the inventors of both extremely large planar LECs andfrozen-junction LECs two key LEC concepts that will be briefly introduced belowbefore the optical scanning experiments are presented in Sects 22 23 and 24
212 Frozen-Junction LECs
Despite possessing some very attractive device characteristics polymer LECs arenot without drawbacks The very operation mechanism responsible for LECrsquosinsensitivity to the active layer thickness and the electrode work function also
2 Optical-Beam-Induced-Current Imaging of Planar Polymer hellip 51
brings serious comprises The turn-on or activation of an LEC typically takesseconds or even minutes as the junction is slowly established by the slow movingions [12 14 45] Unlike the doping of silicon by chemical diffusion the in situelectrochemical doping of an LEC is a room temperature process The mobile ionswhich serve as counter ions to compensate the injected electrons and holes do notbecome part of the polymer chain and remain mobile Once the applied voltage biasis removed the LEC junction will eventually disappear as the doped regions relaxback to the undoped state Therefore LECs are slow to turn-on and exhibit stronghysteresis if they are turned on before reaching a fully relaxed state from a previousoperation [46] LECs also suffer from burn-out if a voltage bias much higher than4 V (for sandwich cells) is applied due to the limited electrochemical stabilitywindow of the electrolyte materials used This means that LECs are not suitable forhigh intensity applications
It is apparent that a fixed LEC junction is desirable since it will retain all theadvantages of LECs while also be fast and stable In an LEC fixing the junctionmeans fixing the counter ion placement Realizing that the ion transportmobility ina polymer electrolyte is strongly temperature dependent Gao et al devised a simplemethod to fix the LEC junction by cooling the cell after the junction formation [4647] When the LEC temperature is below the glass transition temperature (Tg) of thepolymer electrolyte the ions and therefore the LEC junction are immobilized Thefirst demonstration of a frozen LEC junction was on a sandwich cell The cell hadan active layer of poly[2-methoxy-5-(2prime-ethylhexyloxy)-p-phenylene vinylene](MEH-PPV)poly[ethylene oxide] (PEO)lithium triflate (LiTf) blend film sand-wiched between an Indium-Tin-oxide (ITO) electrode and an aluminium electrodeThe cell was activated at room temperature with a fixed positive (ITO biasedpositive) or negative voltage bias to emit strongly Subsequently the cell wascooled to 100 K Two key factors contributed to the success of the frozen junctionFirst the voltage bias was maintained until the cell temperature reached the targettemperature Second the target temperature of 100 K was well below the Tg of PEO(about 208 K) This ensured that the LEC junction was fully stabilized Theresulting ldquofrozen-junctionrdquo LEC exhibit much faster response time (ls) than thesame cell operated at room temperature In a frozen-junction LEC the ions areimmobilized and the device response time is no longer limited by the slow dopingprocess The frozen-junction LECs also exhibit diode-like current versus voltageversus light intensity (I-V-L) characteristics as shown in Fig 22 Significant cur-rent and EL had only been observed under forward bias here defined as a bias withthe same polarity as the applied activation bias Varying the polarity of the acti-vation bias can therefore change the polarity of the frozen junction This behaviouris completely different from both regular LECs which can conduct and emit underboth forward and reverse bias at the same time or the polymer LEDs whosepolarity is fixed
The frozen junction also brings a new functionality to the LEC For the firsttime an activated LEC can operate as a photovoltaic (PV) cell for power genera-tion Figure 23 shows the I-V traces of a frozen-junction cell in dark and under
52 F AlTal and J Gao
illumination The low rectification ratio of the dark I-V curves can be attributed tothe high resistance of the film at low temperature The cell however showed apronounced photovoltaic response in either polarity as shown by the dashedcurves It is remarkable that the same cell can exhibit either a positive or a negativeopen-circuit voltage (VOC) (or short-circuit current ISC) depending on the polarityof the activation bias The photovoltaic response of a frozen-junction LEC just likeits EL is no longer dependent on the electrode work functions Rather the LEC p-n or p-i-n junction determines both the electrical and optical properties of the frozencell The large VOC of minus1 V or +13 V suggests a large built-in potential in the LECjunction Adding an electron-accepting polymer to the LEC blend created a moreefficient frozen-junction polymer photovoltaic cell [48]
Fig 22 Current and light vs voltage (I-V-L) data measured at 100 K (upper curves) after coolingunder +4 V bias The LEC was subsequently heated to 300 K without external bias then biased atminus3 V and cooled (after reaching steady state) to 100 K The I-V-L characteristics (lower curves)were reversed after prebiasing at minus3 V with roughly mirror symmetry relative to 0 V Reprintedwith permission from reference [47] Copyright (1997) American Institute of Physics
Fig 23 The photovoltaicresponse of a frozen sandwichLEC at 100 K the upperpanel shows the responsewhen the cell was activatedusing +4 V The lower panelshows the response after thecell was reheated andactivated using minus3 V thenwas frozen again Reprintedwith permission fromreference [47] Copyright(1997) American Institute ofPhysics
2 Optical-Beam-Induced-Current Imaging of Planar Polymer hellip 53
While it is fairly straightforward to cool an activated LEC to stabilize its junctionin a laboratory eventual application of frozen-junction LECs requires the junctionsonce formed to be frozen at room temperature Progress toward this goal has beenmade by using electrolytes with a high Tg [49 50] The junction is formed atelevated temperatures and the cell is subsequently cooled to room temperatureThe LEC junctions can also be fixed chemically [51ndash56] One method used ion pairmonomers in the LEC blend [51] Upon activation these ions dope the polymer andcause radical-induced polymerization that would significantly reduce the ionsmobility and fix the junction Pei et al used PEO oligomer capped withmethacrylate as the ions conducting component It was found that polymerizing themethacrylate group during junction formation results in a stable junction withlifetime and efficiency comparable to polymer LEDs [57] Another approach uti-lized ionic trapping polymers to establish a permanent junction after junction for-mation [54] Also fixed junctions were formed by incorporating cross-linkablematerials that were cured after the junction formation [58 59] Finally our grouprecently showed that when the ion solvingtransporting material such as PEO wasremoved altogether the resulting cell now only contains a luminescent polymerand a lithium salt could still be activated by applying a much higher bias voltageThe activated cells exhibit characteristics of LECs with strong evidence of dopingMore important the activated state was stable for more than 100 hours without anapplied bias This is the longest reported shelf-life of a frozen junction at roomtemperature [60 61]
213 Extremely Large Planar LECs
Unlike organic or polymer LEDs LECs can operate in both sandwich and planarconfigurations In a planar configuration device the overall device resistance isdominated by the bulk resistance of the active layer which can be enormous for anundoped semiconducting material if the interelectrode spacing is large Forexample a planar polymer LED with an interelectrode spacing of 30 lm exhibitedan EL turn-on voltage of 500 V and was only operational at liquid nitrogen tem-peratures [62] A planar LEC of similar dimensions however can be turned on toemitted light with a mere 4 V bias [8] In an LEC the presence of mobile ions andthe subsequent in situ electrochemical doping render the active film highly con-ductive The images of these planar LECs offered the first visualization of an LECjunction [8 9 63] The relatively small size (interelectrode gap size) of these earlyplanar LECs means that they were difficult to fabricate and to study There had beenalmost no follow-up studies of planar LECs in the late 90s and early 2000s In2003 Gao and Dane demonstrated planar LECs with an interelectrode gap size of15 mm [64] An 800 V bias was applied to turn on two 15 mm cells in series atroom temperature as shown in Fig 24 The millimetre-sized extremely largeplanar LECs are easy to fabricate via shadow masking compared to photolitho-graphic patterning More important the slow turn-on process of these planar LECs
54 F AlTal and J Gao
is highly advantageous for time-resolved studies of the dynamic doping process andthe effects of various operational and material parameters
In situ electrochemical doping of the LEC film affects not only its electricalconductivity but also its optical properties Doping introduces mid-gap impuritystates that quench the PL of the luminescent polymer [65 66] The optical effect ofdoping had been exploited by Gao and Dane to elucidate the very doping process ofLECs [67] Extremely large planar LECs were imaged under UV light illuminationand for the first time the dynamic LEC doping process had been visualizedFigure 25 displays the time-lapse PL images of a 15 mm planar LEC under avoltage bias of 140 V The LEC film exhibits the characteristic orange-red PL ofMEH-PPV Also visible are finger-like darkened regions expanding from the anodetoward the cathode On the cathode side faint but discernible darkening of thepolymer film could be observed These darkened regions are in fact p- and n-doped regions whose PL had been partially quenched The darker p-doped regionexpanded at a faster speed than the n-doped region The expansion stopped once thepropagating doping fronts had met to form a p-n junction Moreover strong ELcould be observed in the last image from the forward-biased p-n junction Thesevisualizations provide indisputable proof that doping did occur in a polymer LECand the formation of a p-n junction was necessary for EL to occur
From the time-lapse images of planar LECs the average doping propagationspeed was extracted and shown to be highly sensitive to the operating temperature[68] By moderately increasing the operating temperature Gao et al successfullydemonstrated the largest planar LEC ever with a gap size of over 10 mm [69]Edman et al on the other hand showed that planar LECs with a gap size of 1 mmcould be turned on with only a 5 V bias when heated to 360 K [70] Figure 26shows an example of the largest planar LEC under UV illumination during theactivation process [71] In this cell both p- and n-doping are clearly visible Also
Fig 24 The photograph oftwo working 15 mmpolymer-based LECs in seriesunder 800 V bias Alsoshown is the deviceconfiguration and biasingcondition The green-emittingdevice is made with a greenemitter and theorange-emitting device ismade with MEH-PPV Theshutter speed is 5 s and theaperture is f10 Reprintedwith permission fromreference [64] Copyright(2003) American Institute ofPhysics
2 Optical-Beam-Induced-Current Imaging of Planar Polymer hellip 55
once again EL was only observed when the p- and n-doping fronts had madecontact to form a p-n junction The cell current had increased by several orders ofmagnitude during the activation process Subsequently the cell was cooled tofreeze the junction The large surface area of this cell allowed for contact probingthe cell surface in a micromanipulated cryogenic probe station
The time-lapse fluorescence imaging of extremely large planar LECs has provento be a powerful and versatile technique in the elucidation of LEC processes Theeffect of thermal annealing [72] electrode work function [73 74] electrolyte salt[75ndash77] and operating voltage [78] had all been studied The static doping profile of
Fig 25 Photographs of a working 15 mm MEH-PPV polymer-based LEC under 365 nm UVillumination The device was tested at 310 K under a voltage bias of 140 V The electrode to theleft is positively biased (anode denoted as ldquo+rdquo) relative to the electrode to the right (cathodedenoted as ldquominusrdquo The photographs were taken at different times after the application of the voltagebias a 8 min b 13 min c 18 min d 43 min The exposure time is 20 s The aperture is f10Reprinted with permission from reference [67] Copyright (2004) American Institute of Physics
56 F AlTal and J Gao
a frozen-junction LEC had been directly observed in a frozen planar LEC [68]Heating the frozen cell briefly however led to partial relaxation of doping and theformation of a p-i-n junction [79] Under the right conditions the frozen p-i-n junction is a much more efficient emitter than an as-formed p-n junction due to theformerrsquos less quenched emission zone A frozen p-i-n junction also exhibits arecord open-circuit voltage [80]
22 Scanning Optical Imaging of Planar LECs
221 The Optical-Beam-Induced Current (OBIC)Technique
Passive time-lapse fluorescence imaging of the entire planar LECs has led to manydiscoveries described above The fully exposed surface of a planar LEC also offers
Fig 26 Time-lapse fluorescence imaging of a 104 mm MEH-PPV PEO CsClO4 planar LECduring turn-on and cooling A fixed DC bias of 400 V was applied to turn on the cell which was at335 K and under UV illumination Time since the DC bias was applied to the cell a no biasb 2 min c 5 min d 8 min e 19 min f 37 min g 54 min Panel H shows the cell current andtemperature as a function of time after the DC bias was applied Uniform enhancement (Leveladjustment in Photoshop) has been applied to images A-G Reprinted with permission fromreference [71] Copyright (2011) American Chemical Society
2 Optical-Beam-Induced-Current Imaging of Planar Polymer hellip 57
a unique platform to perform spatially resolved electrical probing of local electricalproperties of the cell Recently scanning probe microscopic techniques have beenapplied to planar LECs to determine the electric potential distribution across biasedplanar LECs operated at room temperature [81ndash84] With the extremely large104 mm planar cell shown in Fig 26 Gao and Hu used a micromanipulatedcryogenic probe station to map both the electrical potential and the conductivityprofiles of the frozen cell [71] These studies establish that the planar LEC is dopedand a p-n junction is formed The p-doped polymer (MEH-PPV) is much moreconductive than the n-doped polymer and the level of doping is not constant ineach of the doped regions
While the above scanning probe studies aim to map the spatial distribution of anexternally applied potential the LEC junction also possesses a built-inpotentialfield just like a conventional p-n junction The presence of the junctionbuilt-in potentialfield is evidenced by the strong PV response of a sandwichfrozen-junction LEC described in Sect 212 In planar frozen-junction cells VOC
approaching the magnitude of the band gap energy has been observed despite theuse of identical electrodes [80]
For any semiconductor homojunction it is important to know the junctiondepletion width in order to design a more efficient device structure The depletionwidth plays a major role in determining the response time carrier recombinationand photogeneration of the device The optical-beam-induced current (OBIC)technique is especially well suited to probe the electronic structure of a semicon-ductor junction [85ndash89] In OBIC measurement a focused light beam is scannedacross the device a photocurrent is generated when the focused beam illuminatesthe depletion region In neutral p- or n-doped regions by contrast a null OBICsignal is expected due to the absence of a built-in field that can sweep the photo-generated charge carriers before they recombine Figure 27 illustrates a p-n junc-tion under illumination when connected to a load resistor Absorption of photonswith energy larger than the band gap energy generates electron and hole pairs Theelectrons and holes generated in the depletion region are subsequently swept toopposite directions shown This creates a photocurrent and a voltage drop across thejunction The flow of a net current is reflected by the gradient in the Fermi level ofthe junction It should be mentioned that the OBIC technique is typically carried outunder the short-circuit condition where the detected OBIC is a short-circuit currentIn addition photogeneration just outside of the depletion region (within one dif-fusion length) can also give rise to an OBIC signal when the photogenerated chargecarriers enter the depletion region by diffusion
In the absence of an externally applied electric field the drift-diffusion equationsthat govern the current generation in a semiconductor are given by
58 F AlTal and J Gao
Jn=q frac14 kBTq
lnrn lnnru
Jp=q frac14 kBTq
lprp lPpru
) JT=q frac14 kBTq
ethlnrn lprpTHORN ruethlnnthorn lppTHORN
eth21THORN
where J is the current density q is the elementary charge kB is the Boltzmannconstant T is the temperature n is the free electrons concentration p is the holesconcentration l is the mobility coefficient and u is the electrostatic potential Thesubscripts n p and T stand for electrons holes and total respectively The first partof the last line in the equation is the diffusion part of the current density while thesecond part is the drift part The Einstein relation was assumed to be valid Theelectron and hole components of the diffusion current counter each other and vanishwhen the diffusion coefficients of electrons and holes are equal assuming balancedelectronndashhole generationrecombination Also if electrons and holes have differentmobilities the higher mobility component will be more depleted from the gener-ation zone Hence it is expected that the electron and hole diffusion currents tend tocancel each other and minimize the net diffusion current On the other hand
+
+
-
-
R
h
e
EF
EC
EV
Fig 27 A schematic of a p-n junction under lightillumination and theassociated energy banddiagram
2 Optical-Beam-Induced-Current Imaging of Planar Polymer hellip 59
the electronhole drift currents add up to maximize the net drift current and pro-portional to the electrostatic potential gradient Moreover according to OnsagerndashBraun model free carriers generation rate in organic semiconductors is stronglyenhanced by an electric field [90 91] Finally if a symmetric beam is used forexcitation the electron and hole concentration gradients will be equal around thecentre of the beam and nulls the net diffusion current Therefore it is expected thatthe OBIC scan would generate a significant signal only in regions with an elec-trostatic potential gradient ie a built-in electric field In a p-n junction the elec-trostatic potential varies along the depletion region and it causes the OBIC signal tobe significant there and to null in the neural doped regions It is possible to extractthe electrostatic potential profiles from the OBIC profiles However this is not atrivial problem since the relationship is highly nonlinear The details are beyond thistext [92]
Compared to scanning electron beam or scanning Kelvin probe techniques theOBIC method is simple to implement and widely used to characterize semicon-ductor junction structures and to map the minority carrier lifetime defects localresistance and cell uniformity of thin film solar cells Application of the OBICtechnique to an LEC was first reported by Dick et al [63] A focused Argon laserbeam was scanned across an encapsulated planar LEC approximately 20 lm widemounted on a cooled scanning stage VOC rather than ISC was measured to preventrapid dedoping of the activated cell The detection of a peak VOC where the PLintensity showed a large step is consistent with the presence of a p-n junction Thewidth of the junction was estimated to be about 2 lm or 10 of the interelectrodegap Since the activated device was only cooled to 250 K well above the Tg of theelectrolyte the planar LEC was still prone to dedoping The magnitude of the VOCat only a few tens of lV was minuscule compared to the energy gap of theluminescent polymer Since the I-V characteristics of a p-n junction are not linearthe VOC profile would be different from the OBIC profile in width and shape Thechallenges of optically scanning a fully frozen small planar LEC in a cryogenicvacuum chamber (to avoid condensation) meant it was the only study of its kind in15 years since its publication
With the advent of extremely large planar LECs we carried out several OBICstudies of planar LECs that are frozen in a vacuum cryostat The frozen cells had aninterelectrode spacing ranging from 700 lm to 46 mm The extremely large gapsize made it possible to use a variety of opticalcryogenic setups for added func-tionalities The large planar cells were also easier to fabricate using shadowmasking (versus photolithographic) techniques In all these studies the planar cellswere activated at elevated temperatures and cooled to 200 K or below to freeze thejunction In the remainder of this section we briefly introduce the first two OBICstudies carried out in our lab In Sects 13 and 14 we describe in detail the latestOBICPL scans with a focused laser beam
60 F AlTal and J Gao
222 OBIC Scanning of Planar LECswith a Micromanipulated Cryogenic Probe Station
Hu and Gao turned on a 31 mm planar LEC in a micromanipulated cryogenicprobe station (Janis ST-500) under vacuum [41] The planar LEC had a compo-sition of MEH-PPV (10)PEO(10)CsClO4(3) in weight ratio and a pair of AuAlelectrodes deposited on top of the polymer film The micromanipulated probes wereused to make electrical contacts to the electrodes The planar LEC was activatedwith a voltage bias of 300 V at 335 K The cell current reached 3 mA anddecreased when the cell was gradually cooled to 200 K to freeze the junction Theprobe station was equipped with a fibre optical arm The optical fibre was coupledto a 442 nm He-Cd laser beam and manually scanned across the entire interelec-trode gap of the frozen planar LEC OBIC scans were performed in bothshort-circuit and open-circuit conditions along the same path Figure 28 shows thespatial OBIC and VOC profiles of the scans along the path shown Both profileswere very broad due to the large core diameter of the optical fibre used (200 lm)and the jaggedness of the junction The depletion width of the junction thereforewas not resolved The OBIC and VOC peaks however coincided precisely with theposition of the p-n junction shown above the scan profiles The peak VOC was over06 V (versus a few tens of lV of the initial OBIC study) indicating a significantjunction built-in potential
Fig 28 OBIC photocurrentand photovoltage profiles of a31 mm frozen planar LECmeasured in a Janismicromanipulated cryogenicprobe station Top theactivated cell under UVillumination The verticalwhite lines indicate theelectrodepolymer filminterfaces The yellow line andarrow depict the scan pathand direction BottomPhotocurrent andphotovoltage profiles for thescan path shown at the topReprinted from reference[41] Copyright (2011) withpermission from Elsevier
2 Optical-Beam-Induced-Current Imaging of Planar Polymer hellip 61
223 Concerted OBIC and Scanning PL Imaging of PlanarLECs with a Fluorescence Microscope
Subsequently Inayeh et al utilized a low profile microscopy cryostat and afluorescence microscope to perform concerted PL and OBIC scans of planarfrozen-junction LECs [42] A schematic of this setup is shown in Fig 29 Thescanning optical beam was a focused beam of the mercury lamp attached to thefluorescence microscope An octagon-shaped aperture was placed in the opticalpath Moreover the beam was focused to the surface of the planar LEC with a 40objective to a size of about 35 microm in diameter This represented a significantimprovement in scanning resolution compared to the probe station Moreover thesetup allows for a simultaneous recording of the PL intensity of the film with thephotodiode positioned below the cryostat This configuration was made possible by
Fig 29 Top Image of the undoped MEH-PPVPEOKTf planar LEC with a 10 mm interelec-trode gap Also shown are the illumination spots created using a 10 objective lens and a 40objective lens Bottom Schematic illustrating the experimental setup used to perform the OBIC andphotoluminescence scans Blue light (from 448 nm to 497 nm) originating from the mercury lampis focused through a 40 objective lens The light travels through the cryostat window and excitesthe surface of the device Unabsorbed blue light and photoluminescence from the LEC travelthrough the bottom window of the cryostat A 550 nm longpass filter removes the unabsorbed bluelight The photodiode detects the photoluminescence intensity of the LEC film The LEC ismounted in a microscopic cryostat and kept under vacuum Reprinted with permission fromreference [42] Copyright (2012) American Institute of Physics
62 F AlTal and J Gao
the fact that the microscopy cryostat had both a top and a bottom optical windowsand the copper cold finger was partially hollowed out to let the light beam throughThe beam was scanned across the cell by moving the cryostat which was mountedon a motorized scan stage with steps of 10 microm Figure 210 shows the acquiredOBIC and PL profiles from a 1 mm frozen-junction planar LEC Even without thephotograph shown at the top we can easily observe that the OBIC peak is located atthe junction region where the PL intensity underwent a sharp transition Detailedanalysis however revealed that both the PL transition region and the OBIC peakhad a width comparable to the beam width This indicates that the beam was stillnot narrow enough to resolve the depletion width of the frozen junction Furtherimprovement to the scanning resolution was needed
Fig 210 OBIC photocurrent and photoluminescence intensity profiles of the frozen-junction cellshown in Fig 22 an MEH-PPVPEOKTf planar LEC with a 10 mm interelectrode gap turnedon with 25 V and cooled to 200 K Top portion of the cell illuminated under blue light (the fullcell is not illuminated during the OBIC scan) The blue lines depict the area of the cell exposed tolight during the scan and the blue arrow indicates the direction of the scanning optical beamMiddle photocurrent and fluorescence intensity profiles of the scan path shown at top Bottomphotocurrent and differential change in fluorescence intensity profiles of the same scan pathReprinted with permission from reference [42] Copyright (2012) American Institute of Physics
2 Optical-Beam-Induced-Current Imaging of Planar Polymer hellip 63
23 OBIC and Scanning PL Probing of a Frozen Planar p-i-n Junction
231 Introduction
In Sect 22 we introduced the OBIC technique as a valuable tool to probe theelectronic structure of planar LECs Dick et al first demonstrated the feasibility ofmeasuring the spatially resolved VOC and PL intensity across a planar LECalthough the LEC studied was not fully frozen nor was the photocurrent measured[63] The invention of extremely large planar LECs makes the OBIC techniqueextremely attractive due to their ease of fabrication and compatibility with variousscanning apparatus Our grouprsquos first two attempts at scanning extremely largeplanar LECs were successful in that (1) the cells were sufficiently cooled to freezethe LEC junction (2) the OBIC was measured for the first time as well as the VOCThe latter was on the order of the built-in potential expected and (3) the OBIC andPL signals were obtained concurrently The PL trace provides a reliable referencefor determining the position of the OBIC peak relative to the junction position Theresults showed that the peak OBIC and VOC occurred at the junction Thepolymerelectrode interfaces and much of the neutral doped regions were notphotovoltaic-active These studies described briefly in Sects 222 and 223however did not resolve the depletion width of the LEC junction due to thelimitation of the scanning apparatus The beam size in the second study was about35 lm so it was interesting to note that the increase in the gap size of the planar celldid not lead to a wider depletion width that scales with the size of the planar cell Toimprove the scanning resolution the second scanning setup was modified andemployed to scan a unique planar polymer p-i-n junction [43] For the first time thedepletion width of a planar LEC junction has been resolved from a scanning opticalmeasurement This section describes in detail the experimental setup and theobtained results
232 Experimental Details
A planar LEC with a 700 lm gap size was fabricated on a 15 mm 15 mm1 mm sapphire substrate inside a nitrogen-filled gloveboxevaporator systemThe LEC film was spin cast from a cyclohexanone solution of MEH-PPV PEO andpotassium triflate (KTf) with a weight ratio of 10512 and subsequently dried for5 h at 50 degC After that aluminium electrodes with a thickness of 100 nm weredeposited on top of the LEC film under a vacuum of 15 10minus6 torr Thefinished planar LEC had an active area of 8 mm by 700 lm The planar LEC wasloaded into a microscopy cryostat sealed and transferred out of the glovebox fortesting
64 F AlTal and J Gao
The scanning OBICPL apparatus again consists of a Nikon fluorescencemicroscope and a motorized scanning stage onto which the cryostat is mounted Inaddition a Keithley-237 source measurement unit was used to supply the voltagebias and simultaneously measure the device current during both the activation andthe scanning processes A photodiode recorded the PL intensity of the cell duringthe scan and the cell temperature was controlled via a Cryocon 32 controller Themeasurement was controlled with a custom LabView program These arrangementsare similar to the last OBICPL scan The optics of the scanning setup howeverhad been changed Instead of focusing down the beam of the attached mercurylamp the lamp assembly was removed from the microscope A single mode He-Cdlaser (442 nm) was used as the light source Mirrors and a 6511 Galilean tele-scope were used to steer and expand and elevated the laser beam A plano-convexlens with a focal length of 15 cm was used to couple the beam into the micro-scopersquos rear aperture The dichroic mirror inside the microscope directed the laserbeam into a 10 objective and focused it onto the device under test through thecryostat optical window The 2D beam profile was captured using the CCD cameramounted on top of the microscope The 1D intensity profile of the scanning beamwas obtained by integrating the 2D profile perpendicular to the scanning directionThe resulting profile fits well to a Gaussian with a 1e2 waist diameter of 13 lmabout 110 of the size of the focused lamp beam using the same 10 objective
233 Resolving the Depletion Width of a Planar p-i-nJunction
The 700 lm planar LEC was activated by applying a voltage bias of 150 V The celltemperature was kept at 325 K for about 250 s and subsequently increased to 330 Kto speed up the activation processWhen the device current had reached about 13 mA(t = 850 s) the flow of liquid nitrogen was turned on to cool the device at a rate ofminus0365 Ks until the device reached 200 K The device was illuminated with a365 nm wavelength UV lamp during the activation process The time-lapsefluorescence images of the cell are shown in the top part of Fig 211 Like thelarge planar cells shown in Figs 25 and 26 in situ electrochemical doping mani-fested as PL quenching was again highly visible under UV illumination The dopedregions expanded until the doping fronts met and the EL of the cell started to growstronger Cooling led to a dimmer and red-shifted EL at t = 1300 s The last image inthe second row was taken at 200 K with the 150 V bias removed A significantobservation was a bright line between the p- and n-doped regions This is differentfrom the frozen cells shown in Figs 28 and 210 The bright line has stronger PL thanthe neighbouring p- and n-doped regions and is a less quenched quasi-intrinsic regionTherefore the doping profile of this planar cell is that of a p-i-n junction
The focused laser beam was scanned across the frozen planar LEC with a stepsize of 1 lm in a direction that was perpendicular to the p-i-n junction Each scan
2 Optical-Beam-Induced-Current Imaging of Planar Polymer hellip 65
generated a set of raw OBIC and PL profiles Since the scanning laser beam had asignificant size of about 13 lm in diameter deconvolution was necessary to extractthe true OBICPL profiles This was done by fitting the measured PL and OBICpeaks to separate Gaussian functions The Gaussian laser beam was then decon-voluted from these fitting functions The deconvoluted OBIC and PL curves areshown in the lower part of Fig 211 The photocurrent peak has a 1e2 widthof 18 lm The PL peak on the other hand extends over a wider area
Since the exciton diffusion length in PPVs is on the order of nm the width of thedeconvoluted OBIC peak is a good approximation of the width of the depletionregion We note that the PL profile indeed includes a prominent PL peak in thejunction region Also the OBIC peak is displaced by about 12 lm to the rightclose to the ip interface It is reasonable to assume that the peak OBIC position is
Fig 211 Top device snapshots during the turn-on and freezing process The top electrode(anode) was positively biased relative to the lower electrode (cathode) The p-doped region at thetop is visibly darker than the n-doped region in the lower part of the film Only 1100 lm by700 lm of the entire device active area is shown Bottom Deconvoluted photocurrent and PLprofiles The background in the lower panel shows the frozen LEC pictured at zero bias The scanwas done along the central horizontal line in the figure The UV lamp was turned off Reprintedwith permission from reference [43] Copyright (2015) WILEY-VCH Verlag GmbH amp Co
66 F AlTal and J Gao
also the position of the peak built-in field The most significant discovery of thisstudy is the determination of the depletion width of a planar frozen polymerjunction Moreover the LEC junction has the doping profile of a p-i-n junctionManzanares and Heeger first discussed the possibility of a p-i-n junction in apolymer LEC A majority of the doping profiles observed in extremely large planarLECs however have been those of a p-n junction These as-formed p-n junctionscan relax into a p-i-n junction The OBIC profile of a p-n junction as well as apartially relaxed p-n junction is described in the next section
24 High-Resolution OBIC and Scanning PL Imagingof a Frozen Planar Polymer p-n Junction
241 Introduction
The ultimate goal of a scanning optical measurement when applied to planar LECsis to uncover the inner electronic structure of the LEC junction that is responsiblefor all of LECrsquos functionalities as either a light-emitting device or a photovoltaiccell To achieve this goal the scanning resolution must be sufficiently high com-pared to the features to be resolved The scanning setup described in Sect 23produced a scanning beam of about 13 lm in diameter While it was sufficient toresolve the depletion width of a planar p-i-n junction the beam diameter is still toolarge compared to what is obtainable within the diffraction limit We note that in thefirst OBIC study by Dick et al the laser beam was focused to have a diameter ofabout 1 lm with the use of a 40 microscope objective [63] A high-poweredobjective however could not be used in the previous setup as it led to a secondfocused spot on the device surface This was likely caused by some internalreflections of the laser beam which entered the microscope from the lamp port Inthis section we describe a revised optical scanning setup that has overcome thisproblem The focused beam had a Gaussian shape with a 1e2 width of only 19 micromThe planar LEC itself had also been optimized by fine-tuning the LEC blend theelectrode material and the operating conditions The planar LEC was activated toexhibit a straight long and highly emissive p-n junction that had been successfullyfrozen The concerted high-resolution OBIC and scanning PL imaging the frozenplanar LEC exposed the narrowest p-n junction in a frozen LEC ever reported [44]
242 Experimental Details
The planar LECs in this study had a composition of MEH-PPV (10) PEO (5) KTf(12) by weight and a thickness of about 05 lm A pair of thermally evaporatedgold electrodes defined a 700 microm interelectrode spacing and a cell length of 7 mm
2 Optical-Beam-Induced-Current Imaging of Planar Polymer hellip 67
The same microscopy cryostat was used to house the cell and the device fabricationand testing procedures are identical to those described in Sect 232 The lightsource and optics have been modified The experimental setup is shown inFig 212 A blue single mode diode laser (473 nm) was used as the excitationsource The laser beam was steered using mirrors (M1 M2) and coupled into asingle mode fibre (SMF) via a coupler lens Another lens was attached at the otherterminal of the fibre to collimate the output beam which was then expanded in sizeA 5050 beam splitter cube then redirected the beam into a 50 objective and at thesame time allowed light collected by the objective to pass through This arrange-ment allowed the excitation beam to by-pass the microscope optics In addition themercury lamp was reattached which was used to take fluorescence images of thecell during activation The focused Gaussian beam had a 1e2 width of 19 microm at thedevice surface A silicon photodiode positioned under the bottom optical windowcollected the PL from the LEC film The excitation beam was rejected by alow-pass filter (LPF) with a cut-off wavelength of 500 nm between the opticalwindow and the photodiode The device PL pictures were captured via a 10objective using a CCD camera that is mounted on top of the microscope Thisprovided a field of view that covers an area of 09 mm 18 mm
243 Results and Discussion
The planar LEC was activated by applying a 20 V DC bias at a temperature of360 K Subsequently it was cooled at a rate of 019 Ks to 170 K The lowerfreezing temperature of 170 K compared to 200 K used in previous scans ensured afixed junction for multiple optical scans The frozen junction was verified byrepeated I-V scans which did not show change over the course of the experimentThe top panel of Fig 213 depicts the cell geometry and the time evolution of cellcurrent and cell temperature during the activation process We note that a very high
Fig 212 Left Experimental setup for scanning OBIC and PL imaging of frozen planar LECsRight 2D beam profile and Gaussian fit of the focused excitation beam Reprinted with permissionfrom reference [44] Copyright (2016) American Institute of Physics
68 F AlTal and J Gao
peak current of over 5 mA had been reached which is an indication of strongdoping and good film quality The lower panel displays the time-lapse fluorescenceimages of the cell The LEC film is uniformly fluorescent before the voltage biaswas applied (t = 0 s) The doping fronts of this cell propagated much faster than theearlier cells shown in Sects 22 and 23 At t = 3 s a large part of the LEC film wasalready doped to either p-type on the anode (+) side or n-type on the cathode (minus)side At t = 9 s a continuous light-emitting p-n junction was visible between thedoped regions The p-n junction was initially very uneven but became muchstraighter with time Meanwhile the doped regions became darker indicating anincrease in doping level The darkening of the LEC film was accompanied by therapid increase in cell current The applied voltage bias was removed once the targettemperature of 170 K had been reached The PL image of the frozen cell revealed ap-n junction doping profile A thin bright line was observed in the junction regionHowever the intensity of the bright line was not high enough to suggest a p-i-n junction structure Moreover it will be shown that the as-activated p-n junctioncould be relaxed into a p-i-n junction when subjected to warmingre-cooling cycles
A total of 18 OBICPL scans were performed across the frozen p-n junction asseven locations Figure 214 shows the OBIC and PL profiles of a full
Fig 213 Device activation process Top Time evolution of the cell current and temperatureduring the activation process A DC voltage bias of 20 V was applied The inset shows a schematicof the planar LEC Bottom Time-lapse fluorescence images of the planar LEC during the activationprocess Only a section of the entire cell is shown The bright line formed after t = 9 s is due toelectroluminescence of the cell Reprinted with permission from reference [44] Copyright (2016)American Institute of Physics
2 Optical-Beam-Induced-Current Imaging of Planar Polymer hellip 69
(electrode-to-electrode) scan The OBIC peak was assigned a beam position of zerolm as a reference The PL profile indicates an n-doped region with the highestoverall PL intensity and a darker p-doped region to the right of the OBIC peakA small local PL peak just to the left of the OBIC peak can be attributed to the thinbright line observed in the PL image The sensitivity of the PL scan is evident in thedetection of a PL ldquotransition zonerdquo on the p side of the junction between 0 and +30lm In this zone the PL intensity decreased continuously from the high level of then side to the much lower level of the p side The large variation in PL intensity isalso visible near both electrode interfaces This variation indicates a large dopinggradient and the level of doping is the highest just inside the electrode edges forboth p- and n-doped films This observation is consistent with the results of thecontact probing measurement
The most significant observation is the extremely narrow and prominent OBICpeak at x = 0 lm The lower panel of the figure provides an expanded view of theOBIC peak that incorporated the data points from a total of four scans across thesame junction region The data points of the individual scan were adjustedshiftedso that their peak positions coincide This procedure was necessary to eliminate anyoffset due to the mechanical hysteresis of the scanning stage The data are fitted to aGaussian function from which a 1e2 width of 31 lm was determined Since boththe scanning beam and the measured OBIC profile are Gaussian functionsdeconvolution also results in a Gaussian whose width is given by
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
W2OBIC W2
Beam
p
where W is the 1e2 Gaussian width The deconvoluted OBICpeak width is 245 lm We note that all 18 scans yielded an OBIC peak width thatwas larger than the excitation beam diameter of 19 lm An average of all 18 scansgives an OBIC width of 253 lm with a standard deviation of 045 lm Theaverage OBIC peak width after deconvolution is 15 lm The average OBIC signalis 292 plusmn40 pA for an excitation beam intensity of 2 microW
Fig 214 OBIC and PLprofiles of the frozen planarLEC across the same junctionregion Top Full OBIC andPL scans of the planar LECacross the entire planar LECThe cathode is located atminus230 microm and the anode islocated at 470 microm asindicated by the verticaldashed lines Bottom OBICdata of four scans andGaussian fit to the data nearthe OBIC peak Reprintedwith permission fromreference [44] Copyright(2016) American Institute ofPhysics
70 F AlTal and J Gao
Since the exciton diffusion length in MEH-PPV is on the order of nm the OBICpeak width should represent the junction depletion width The average junctionwidth of 15 lm is the smallest ever measured in a planar LEC regardless of theinterelectrode spacing The junction width accounts for a mere 021 of theinterelectrode gap which is the lowest value reported for LECs The OBIC junctionwidth is also much smaller than the width of the EL zone (10 microm) and the PLtransition zone (30 microm) The small junction width obtained attests to the highresolution of the scans
The drop-off in PL intensity to the right of the OBIC peak indicates a dopinggradient at the edge of the p-n junction There is a strong possibility that the dopinggradient is caused by the presence of rough features on the sub-micrometre scaleSharp protrusions on a larger scale had been observed along a planar p-n junction[80] They contributed to a tunnelling leakage current that degraded the rectificationratio of the as-formed frozen p-n junction The sharp features however could beremoved by subjecting the frozen p-n junction to a warmingcooling cycle thatcaused partial dedoping and smoothing the p-n junction The ldquorelaxedrdquo p-n junctionexhibit improved rectification and a much larger VOC when illuminated
In an attempt to smooth out the junction and eliminate any possible submicronprotrusions the frozen cell was partially relaxed after the aforementioned scans Therelaxationdedoping cycle was carried out by briefly (for a few minutes at a time)warming the frozen cell to 260 K and cooling it back to 170 K During thesethermal cycles the cell was kept at an open-circuit condition to avoid fast dedopingand the loss of the active junction In total four relaxation cycles were carried outThe cell current decreased after each cycle and the I-V curves became morenonlinear and less symmetric
This is a strong indication that dedoping had occurred OBIC scans were per-formed after each dedoping cycle along the same junction location As the cellbecame more resistive the input optical power was increased to obtain a measur-able OBIC signal Figure 215 compare the normalized OBIC profiles withoutdedoping and after the final dedoping cycle The measured OBIC profile narrowedto 2 lm after dedoping After subtracting the beam width the junction width isonly 06 lm compared to 245 lm without dedoping A very prominent PL peakappeared to the left of the OBIC peak The PL profile became similar to that of a p-i-n junction shown in Fig 211 The p-i-n junction formed by controlled dedopinghowever is quite different from an as-formed p-i-n junction discussed in Sect 13Here the removal of the fine features along the edge of the junction resulted in thenarrowest OBIC peak ever reported for a planar LEC The p-i-n junction had sharpboundaries between the differently doped regions Indeed the PL transition regionbetween the ldquoirdquo region and the p region has narrowed to about 3 lm from nearly30 lm in the as-formed p-n junction The more abrupt PL transition is consistentwith a narrowed OBIC profile and a reduced junction width
2 Optical-Beam-Induced-Current Imaging of Planar Polymer hellip 71
25 Conclusion and Outlook
Polymer-based LECs are intriguing and promising devices that offer attractivedevice characteristics not easily attained by other organic electronic devicesDoping plays an essential role in the operation of polymer-based LECs On onehand doping leads to conductivity increase that gives LECs their desirable elec-trical properties On the other hand the optical effect of doping allows for thevisualization of the dynamic doping process via time-lapse fluorescence imaging Inthis chapter we summarized our recent experimental work on the scanning opticalimaging of planar LECs The experiments exploited both the electrical and opticaleffects of doping as well as the temperature dependence of ionic conductivity of theLEC film For the first time we had resolved the junction depletion width of aplanar p-i-n junction and a planar p-n junction The narrowing of the junction uponthermal cycling strongly suggests the presence of fine structures on the edge of thejunction For an as-formed frozen p-n junction the depletion width accounted foronly 02 of the entire cell area enclosed by the electrodes Since only the junctionregion contributes to EL and PV response in an LEC the narrow junction width isnot ideal for efficient operation of LECs as either a light-emitting device or a PVcell An igneous solution to this problem is to form multiple junctions that aresimultaneously emitting without increasing the total cell area By introducing dis-persed metallic particles to the LEC film we successfully demonstrated a newdevice structure called a bulk homojunction (BHoJ) LEC [93ndash96] A BHoJ planarLEC exhibits vastly improved effective emitting area as well as a giant VOC whenoperated as a PV cell BHoJ LECs in a sandwich configuration and frozen at roomtemperature represent a major challenge and opportunity in LEC research
Acknowledgements The optical scanning studies were supported by the Natural Sciences andEngineering Research Council of Canada (NSERC) Faleh AlTal is supported by an OntarioTrillium Scholarship
Fig 215 Normalized OBICscans at the same locationbefore and after four dedopingcycles The excitation laserpower used was 2 microW (red)before dedoping and 50 microW(blue) after four dedopingcycles Reprinted withpermission from reference[44] Copyright (2016)American Institute of Physics
72 F AlTal and J Gao
References
1 SB Meier D Tordera A Pertegas C Roldan-Carmona E Orti HJ Bolink Mater Today17 217ndash223 (2014)
2 Y Xiong L Li JJ Liang H Gao SY Chou QB Pei Mater Horiz 2 338ndash343 (2015)3 A Sandstrom L Edman Energy Technol 3 329ndash339 (2015)4 Q Pei Y Yang G Yu Y Cao AJ Heeger Synth Met 85 1229ndash1232 (1997)5 Q Sun Y Li Q Pei J Disp Technol 3 211ndash224 (2007)6 JF Fang YL Yang L Edman Appl Phys Lett 93 (2008)7 JM Leger Adv Mater 20 837ndash841 (2008)8 QB Pei G Yu C Zhang Y Yang AJ Heeger Science 269 1086ndash1088 (1995)9 QB Pei Y Yang G Yu C Zhang AJ Heeger J Am Chem Soc 118 3922ndash3929 (1996)
10 RH Friend RW Gymer AB Holmes JH Burroughes RN Marks C Taliani DDCBradley DA Dos Santos JL Bredas M Logdlund WR Salaneck Nature 397 121ndash128(1999)
11 Y Cao ID Parker G Yu C Zhang AJ Heeger Nature 397 414ndash417 (1999)12 Y Cao QB Pei MR Andersson G Yu AJ Heeger J Electrochem Soc 144 L317ndashL320
(1997)13 Y Cao G Yu AJ Heeger CY Yang Appl Phys Lett 68 3218ndash3220 (1996)14 Y Yang QB Pei J Appl Phys 81 3294ndash3298 (1997)15 ZT Zhang KP Guo YM Li XY Li GZ Guan HP Li YF Luo FY Zhao Q Zhang
B Wei QB Pei HS Peng Nat Photonics 9 233ndash238 (2015)16 Y Nishikitani D Takizawa H Nishide S Uchida S Nishimura J Phys Chem C 119
28701ndash28710 (2015)17 KO Burnett PP Crooker NM Haegel Y Yoshioka D MacKenzie Synth Met 161
1496ndash1499 (2011)18 ZZ You GJ Hua Int J Electron 97 99ndash104 (2010)19 J Mindemark S Tang J Wang N Kaihovirta D Brandell L Edman Chem Mater 28
2618ndash2623 (2016)20 J Mindemark L Edman J Mater Chem C 4 420ndash432 (2016)21 Z Shu O Pabst E Beckert R Eberhardt A Tunnermann Mater Today Proc 3 733ndash738
(2016)22 JH Jang LH Kim YJ Jeong K Kim TK An SH Kim CE Park Org Electron 34
50ndash56 (2016)23 A Asadpoordarvish A Sandstrom L Edman Adv Eng Mater 18 105ndash110 (2016)24 K Shanmugasundaram MS Subeesh CD Sunesh Y Choe RSC Adv 6 28912ndash28918
(2016)25 A Pertegas MY Wong M Sessolo E Zysman-Colman HJ Bolink Ecs J Sol State Sci
Techn 5 R3160ndashR3163 (2016)26 QY Zeng FS Li TL Guo GG Shan ZM Su Sci Rep 6 9 (2016)27 JQ Wu FS Li QY Zeng C Nie PC Ooi TL Guo GG Shan ZM Su Org Electron
28 314ndash318 (2016)28 MS Subeesh K Shanmugasundaram CD Sunesh RK Chitumalla J Fang Y Choe
J Phys Chem C 120 12207ndash12217 (2016)29 DAW Ross PA Scattergood A Babaei A Pertegas HJ Bolink PIP Elliott Dalton
Transact 45 7748ndash7757 (2016)30 LD Bastatas KY Lin MD Moore KJ Suhr MH Bowler YL Shen BJ Holliday JD
Slinker Langmuir 32 9468ndash9474 (2016)31 MD Weber M Adam RR Tykwinski RD Costa Adv Funct Mater 25 5066ndash5074
(2015)32 MF Ayguler MD Weber BMD Puscher DD Medina P Docampo RD Costa J Phys
Chem C 119 12047ndash12054 (2015)33 Q Pei AJ Heeger Nat Mater 7 167 (2008)
2 Optical-Beam-Induced-Current Imaging of Planar Polymer hellip 73
34 GG Malliaras JD Slinker JA DeFranco MJ Jaquith WR Silveira Y-W Zhong JMMoran-Mirabal HG Craighead HD Abruna JA Marohn Nat Mater 7 168 (2008)
35 S Jenatsch L Wang M Bulloni AC Veron B Ruhstaller S Altazin F Nuesch R HanyACS Appl Mater Interf 8 6554ndash6562 (2016)
36 S van Reenen RAJ Janssen M Kemerink Adv Funct Mater 25 3066ndash3073 (2015)37 S van Reenen P Matyba A Dzwilewski RAJ Janssen L Edman M Kemerink J Am
Chem Soc 132 13776ndash13781 (2010)38 S van Reenen RAJ Janssen M Kemerink Org Electron 12 1746ndash1753 (2011)39 G Gozzi LF Santos RM Faria Epl 100 (2012)40 A Munar A Sandstrom S Tang L Edman Adv Funct Mater 22 1511ndash1517 (2012)41 YF Hu B Dorin F Teng J Gao Org Electron 13 361ndash365 (2012)42 A Inayeh B Dorin J Gao Appl Phys Lett 101 253305 (2012)43 F AlTal J Gao Phys Status Solidi Rapid Res Lett 9 77ndash81 (2015)44 F AlTal J Gao J Appl Phys 120 8 (2016)45 Y Shao GC Bazan AJ Heeger Adv Mater 19 365ndash370 (2007)46 J Gao YF Li G Yu AJ Heeger J Appl Phys 86 4594ndash4599 (1999)47 J Gao G Yu AJ Heeger Appl Phys Lett 71 1293ndash1295 (1997)48 J Gao G Yu AJ Heeger Adv Mater 10 692ndash695 (1998)49 G Wantz B Gautier F Dumur TNT Phan D Gigmes L Hirsch J Gao Org Electron
13 1859ndash1864 (2012)50 G Yu Y Cao M Andersson J Gao AJ Heeger Adv Mater 10 385ndash388 (1998)51 JM Leger DB Rodovsky GR Bartholomew Adv Mater 18 3130 (2006)52 DT Simon DB Stanislowski SA Carter Appl Phys Lett 90 103508 (2007)53 JM Leger DG Patel DB Rodovsky GP Bartholomew Adv Funct Mater 18 1212ndash
1219 (2008)54 CV Hoven HP Wang M Elbing L Garner D Winkelhaus GC Bazan Nat Mater 9
249ndash252 (2010)55 IV Kosilkin MS Martens MP Murphy JM Leger Chem Mater 22 4838ndash4840 (2010)56 S Tang K Irgum L Edman Org Electron 11 1079ndash1087 (2010)57 ZB Yu L Li HE Gao QB Pei Sci China Chem 56 1075ndash1086 (2013)58 ZB Yu ML Sun QB Pei J Phys Chem B 113 8481ndash8486 (2009)59 S Tang L Edman Electrochim Acta 56 10473ndash10478 (2011)60 B Gautier J Gao Appl Phys Lett 101 093302 (2012)61 B Gautier XM Wu F Altal SL Chen J Gao Org Electron 28 47ndash52 (2016)62 U Lemmer D Vacar D Moses AJ Heeger T Ohnishi T Noguchi Appl Phys Lett 68
3007ndash3009 (1996)63 DJ Dick AJ Heeger Y Yang QB Pei Adv Mater 8 985ndash987 (1996)64 J Gao J Dane Appl Phys Lett 83 3027ndash3029 (2003)65 AL Holt JM Leger SA Carter J Chem Phys 123 044704 (2005)66 N Kaihovirta A Asadpoordarvish A Sandstrom L Edman ACS Photonics 1 182ndash189
(2014)67 J Gao J Dane Appl Phys Lett 84 2778ndash2780 (2004)68 J Dane C Tracy J Gao Appl Phys Lett 86 153509 (2005)69 YF Hu C Tracy J Gao Appl Phys Lett 88 123507 (2006)70 JH Shin A Dzwilewski A Iwasiewicz S Xiao A Fransson GN Ankah L Edman
App Phys Let 89 013509 (2006)71 YF Hu J Gao J Am Chem Soc 133 2227ndash2231 (2011)72 S Alem J Gao Org Electron 9 347ndash354 (2008)73 D Hohertz J Gao Adv Mater 20 3298ndash3302 (2008)74 JH Shin P Matyba ND Robinson L Edman Electrochim Acta 52 6456ndash6462 (2007)75 YF Hu J Gao Appl Phys Lett 89 253514 (2006)76 Y Hu Y Zhang J Gao Adv Mater 18 2880 (2006)77 J-H Shin ND Robinson S Xiao L Edman Adv Funct Mater 17 1807ndash1813 (2007)78 ND Robinson J-H Shin M Berggren L Edman Phys Rev B 74 155210 (2006)
74 F AlTal and J Gao
79 YG Zhang YF Hu J Gao Appl Phys Lett 88 163507 (2006)80 Y Zhang Y Hu J Gao Appl Phys Lett 91 233509 (2007)81 P Matyba K Maturova M Kemerink ND Robinson L Edman Nat Mater 8 672ndash676
(2009)82 LSC Pingree DB Rodovsky DC Coffey GP Bartholomew DS Ginger J Am Chem
Soc 129 15903ndash15910 (2007)83 DB Rodovsky OG Reid LSC Pingree DS Ginger ACS Nano 4 2673ndash2680 (2010)84 JD Slinker JA DeFranco MJ Jaquith WR Silveira YW Zhong JM Moran-Mirabal
HG Craighead HD Abruna JA Marohn GG Malliaras Nat Mater 6 894ndash899 (2007)85 T Wilson C Sheppard Theory and Practice of Scanning Optical Microscopy (Academic
Press London Orlando 1984)86 M Ogura K Sakaue Y Tokumaru Jap J Appl Phys Part 2 Lett 24 L617ndashL619 (1985)87 C Werner D Reuter AD Wieck Phys E Low Dimens Syst Nanostruct 32 508ndash511
(2006)88 MG Liu SL Wang JC Campbell JD Beck CF Wan MA Kinch J Appl Phys 98
(2005)89 C Raynaud DM Nguyen N Dheilly D Tournier P Brosselard M Lazar D Planson
Phys Status Solidi A Appl Mat 206 2273ndash2283 (2009)90 L Onsager Phys Rev 54 554ndash557 (1938)91 CL Braun J Chem Phys 80 4157ndash4161 (1984)92 S Busenberg WF Fang K Ito Siam J Appl Math 53 187ndash204 (1993)93 C Tracy J Gao Appl Phys Lett 87 143502 (2005)94 C Tracy J Gao J Appl Phys 100 104503 (2006)95 W Bonnet C Tracy G Wantz GJ Liu J Gao Small 4 1707ndash1710 (2008)96 SL Chen G Wantz L Bouffier J Gao Chem Electron Chem 3 392ndash398 (2016)
2 Optical-Beam-Induced-Current Imaging of Planar Polymer hellip 75
Chapter 3Optical Engineering of Light-EmittingElectrochemical Cells IncludingMicrocavity Effect and OutcouplingExtraction Technologies
Hai-Ching Su
Abstract Light-emitting electrochemical cells (LEC) have received much scientificinterest after the first demonstration in 1995 In addition to enormous progresses indevelopment of novel emissive materials for LECs charge carrier balance has beenimproved to significantly enhance device efficiency of LECs However furtherimprovement in device performance is still required to meet industrial applicationsOptical engineering techniques involving interference scattering and waveguidingeffects in device optical structures would be feasible approaches to modify theelectroluminescence (EL) spectrum and to extract more light output from LECs Inthis chapter the microcavity effect and outcoupling extraction technologies arereviewed in detail This complements the introduction provided in Chap 1Microcavity effect can be employed to tailor the EL spectrum of LECs by adjustingthe interference effect Scattering from microlens array increases light outcouplingfrom substrates of LECs Furthermore color conversion and waveguide couplingresult in more light extraction since both EL trapped in waveguide and substratemode would be harvested With these promising techniques doubled device effi-ciency can be realized These results confirm that optical engineering techniques arenecessary to achieve highly efficient LECs
Keywords Light outcoupling Waveguiding Microcavity Color conversionlayer
H-C Su (amp)Institute of Lighting and Energy Photonics National Chiao Tung UniversityTainan 71150 Taiwane-mail haichingsumailnctuedutw
copy Springer International Publishing AG 2017RD Costa (ed) Light-Emitting Electrochemical CellsDOI 101007978-3-319-58613-7_3
77
31 Introduction
311 Microcavity Effect in Organic Thin-Film Devices
As explained in Chap 1 LEC architecture involves a single active layer of around100ndash200 nm sandwiched in between two electrodes Since the active layer thick-ness is generally similar to the visible light wavelength and a high-reflectance metalis commonly employed as the cathode (eg Al Ag etc) the wavelength depen-dent emission characteristics of the active layer are adjusted in a microcavitystructure due to optical interference This modifies the density of optical mode andchanges the electroluminescence (EL) spectrum spectrally Output spectrum from abottom emitting thin-film light-emitting device would be simulated by employingthe equation shown in the following [1]
EextethkTHORNj j2frac14T2 1
N
PN
ifrac1411thornR1 thorn 2
ffiffiffiffiffiR1
pcos 4pzi
k thornu1
1thornR1R2 2ffiffiffiffiffiffiffiffiffiffiR1R2
pcos 4pL
k thornu1 thornu2
EintethkTHORNj j2 eth31THORN
where R1 and R2 are the reflectances from cathode and substrate respectively u1
and u2 are the phase changes of reflection from cathode and substrate respectivelyT2 is the transmittance from the substrate L is the total optical thickness of theactive layers EintethkTHORNj j2 is the emission spectrum without microcavity effectEextethkTHORNj j2 is the emission spectrum coming out from the substrate zi is the opticaldistance between cathode and the emitting sublayer i The emitting zone is parti-tioned into N sublayers and all contributions are integrated The photoluminescence(PL) spectrum of the emissive material on a quartz substrate is used as the emissionspectrum without microcavity effect because no high-reflectance metal is present
It is noted that the optical structure of the device eg active layer thickness andelectrode structure would affect the output EL significantly Furthermore therecombination zone of LECs is moving while the doped layers are extendingresulting in temporal evolution in EL spectrum [2] Hence the EL spectrum ofLECs is modified by tuning the active layer thickness andor the recombinationzone position
312 Optical Modes in Organic Thin-Film Devices
In a typical thin-film organic-based device most of the EL in the active layer istrapped in indium tin oxide (ITO) and organic layer because of a large discrepancyin the refractive indices between the high-index layers (norg 17 and nITO 19)and the low-index glass substrate (nglass 15) The percentage of EL in waveguidemode is 40ndash60 of the total intensity [3] Another emission part is trapped in the
78 H-C Su
glass substrate because of total internal reflection at the output end Eventually only20 of total intensity can be harvested in the forward direction (external mode)Schematic diagram of the optical modes in an organic thin-film device is shown inFig 31 Hence recycling the EL in waveguide and substrate mode is a possibleway to enhance device performance of LECs
313 Organization of This Chapter
In the following the state of the art on optical engineering of LECs includingmicrocavity effect and outcoupling extraction technologies are reviewed Themicrocavity effect can be used to tailor the output EL of LECs Blue-green emissionform white LECs is suppressed to improve the color purity White EL can begenerated by constructive interference from LECs based on a single blue-emittingmaterial In addition near-infrared EL can also be obtained by constructive inter-ference from LECs based on deep-red-emitting material Such technique enhancesobtainable EL spectral range from a limited number of available emissive materialsfor LECs Outcoupling extraction technologies are essential to increase deviceefficiency of LECs Recycling the confined EL by employing red color conversionlayers and waveguide coupling are mentioned Finally conclusion and outlook arediscussed
32 Tailoring Output EL Spectrum of LECsby Employing Microcavity Effect
321 Suppression of Blue-Green Emissionto Achieve Purer White EL
Modifying EL spectra by adjusting microcavity effect is useful for white LECsbased on ionic transition-metal complexes (iTMCs) to achieve white EL with
Substrate
ITO anode
Organic layers
Metal cathode
External mode
Substrate mode
Waveguide mode
Fig 31 Schematic diagramof the optical modes in anorganic thin-film device
3 Optical Engineering of Light-Emitting Electrochemical Cells hellip 79
Commission Internationale de lprimeEclairage (CIE) coordinates near (033 033)because efficient saturated blue-emitting iTMCs were scarce Published white LECscontaining sky-blue-emitting iTMCs commonly show greenish white emission evenmixed with deep-red-emitting iTMCs [4ndash7] As shown in the work published by Suet al [8] reducing green emission part of sky-blue-emitting iTMCs by destructiveinterference is a possible approach to reach saturated blue EL and purer white ELemission was obtained consequently The blue-green-emitting iTMC [Ir(df-ppz)2(dtb-bpy)][PF6] (1) (where dfppz is 1-(24-difluorophenyl)pyrazole anddtb-bpy is [44prime-di(tert-butyl)-22prime-bipyridine]) proposed by Tamayo et al wasemployed as the host complex [9] Efficient red laser dye Sulforhodamine 101 (2)which was shown to have a high photoluminescence quantum yield (PLQY) of095 plusmn 002 [10] was used as the guest dopant The calculated and measuredoutput EL spectra of the white LEC based on an active layer (320 nm) doped with03wt 2 are shown in the left part of Fig 32 When the emission zone of thewhite LECs was placed at 195 nm away from cathode the calculated and measuredoutput EL spectra were well fitted This emitting zone position was rational becausea larger energy level offset in the highest occupied molecular orbital (HOMO) levelbetween host and guest (Fig 33) lead to more pronounced hole trapping and therecombination zone was thus nearer the anode Reduced bandwidth of the blue ELwas attributed to destructive interference of the green emission from complex 1Saturated blue EL lead to purer white emission with CIE coordinates of (032030) which was close to equal energy point (033 033) When the active layerthickness was changed the spectral region at which destructive interferenceoccurred was different and the output EL emission was modified As shown in theright part of Fig 32 destructive interference took place at the red spectral regionwhen the emissive layer thickness was 270 nm Calculation also predictedwell-fitted spectrum with the experimental data (right part of Fig 32) These resultsshowed that microcavity effect can be employed to modify the output EL emissionand purer white EL can be obtained without using deep-blue emitting materials
00
02
04
06
08
10 Measured Simulated
EL In
tens
ity (a
u)
Wavelength (nm)400 500 600 700 400 500 600 700
00
02
04
06
08
10 Measured Simulated
EL In
tens
ity (a
u)
Wavelength (nm)
Fig 32 Calculated and measured EL spectra from the white LECs based on emissive layerthicknesses of 320 nm (left) and 270 nm (right)
80 H-C Su
322 Non-doped White LECs Based on a SingleEmissive Material
White LECs have recently attracted much attention since they are potential insolid-state lighting applications White iTMC LECs were commonly composed ofhost-guest active layers to reach better device performance Nevertheless lowdoping concentrations (lt1 wt) of the red-emitting guest in the blue-emitting hostlayer complicated the process steps and run-to-run doping concentration variationlead to color migration for each device In a recently published work non-dopedwhite LECs based on microcavity effect were demonstrated [11] The blue-emittingiTMC (1) was used as the host material in the emissive layer If a thicker activelayer (490 nm) was used red EL was increased by microcavity effect when therecombination zone shifted to certain locations Therefore white EL was obtainedby mixing blue-green EL from 1 and enhanced red EL from microcavity effect
The fitted EL spectra and then estimated recombination zone positions at dif-ferent times are depicted in Fig 34 Initially the LECs based on 1 showedblue-green EL with a peak at 490 nm The EL then exhibited blue shift and the fullwidth at the half maximum (FWHM) decreased gradually (the left part of Fig 34)After ca 1 h a red shoulder increased gradually and the EL was getting white (thecentral part of Fig 34) Eventually the red emission reduced and the EL turnedinto blue-green again (the right part of Fig 34) The EL turned into white at 75ndash90 min while the EL was greenish blue outside this period This temporal ELspectra may not possibly be related to material degradation because almostchanged EL spectrum was measured in thinner devices (ca 200 nm) containingcomplex 1 [12 13] Similar results were measured in LECs based on thicker(gt450 nm) emissive layers of ruthenium(II) complexes [2 14 15] and thetime-dependent EL spectrum resulted from modified microcavity effect due torecombination zone moving [2]
To clarify the mechanism of the time-dependent EL spectrum from the LECscontaining complex 1 device physics responsible for spectral changing areexplained in the following The recombination zone was initially at the centralemissive layer (zi = 300ndash230 nm at 15ndash57 min the left part of Fig 34)Blue-green constructive interference occurred and thus the EL showed blue shiftand narrowed FWHM Then the recombination zone shifted towards the cathode
Fig 33 Energy leveldiagram of the host (1) andthe guest (2) molecules
3 Optical Engineering of Light-Emitting Electrochemical Cells hellip 81
with time (zi = 206ndash190 nm at 66ndash81 min the central part of Fig 34) Theblue-green emission showed more blue-shifted and even narrower FWHM Inaddition another constructive interference took place at red and dual-emissionwhite emission was obtained When the recombination zone was further moving tothe cathode (zi = 170ndash120 nm at 90ndash174 min the right part of Fig 34) the redemission reduced and finally disappeared After ca 3 h the EL was reaching asteady state which revealed that stabilized recombination zone was present becauseof well formed doped layers Since the time-dependent EL spectrum was matchedby the simulation from microcavity effect when the recombination zone wasmoving EL spectral migration resulted from material degradation may be ruled outTime-dependent recombination zone was rationalized by the energy level for theLECs containing complex 1 (Fig 35) When a bias voltage was applied therecombination zone was at the central active layer in spite of a lower hole injectionbarrier (085 eV) than electron injection barrier (129 eV) It may be related tohigher electron mobility in a thicker film (490 nm) of complex 1 at 37 V Due tothe difference in hole and electron injection barrier the necessary number of ions atthe anode to reach ohmic contact for hole is much lesser than that for electron at thecathode The temporal increasing rate for hole injection efficiency is thus higherthan that for electrons leading to a recombination zone moving from the center ofthe emissive layer to the cathode After the doped layers were completely estab-lished the p-i-n layer was steady and the recombination zone froze These resultsshowed that recombination zone moving lead to red constructive interferencewhich combined with blue-green emission from complex 1 to reach white emissionIt demonstrated a simple approach to realize white EL from non-doped LECs
00
02
04
06
08
10 t (min) Measured z (nm)15 30030 28057 230
EL In
tens
ity (a
u)
Wavelength (nm)
00
02
04
06
08
10 t (min) Measured z (nm)66 20675 19678 19381 190
EL In
tens
ity (a
u)
Wavelength (nm)400 500 600 700 800 400 500 600 700 800 400 500 600 700 800
00
02
04
06
08
10 t (min) Measured z (nm)90 170
105 150174 120
EL In
tens
ity (a
u)
Wavelength (nm)
Fig 34 Measured and simulated temporal EL spectra from the white LECs containing 1 at 15ndash57(left) 66ndash81 (central) and 90ndash174 min (right) under 37 V The temporal recombination zoneposition measured from cathode was extracted by calculations and is labeled for comparison
Fig 35 The energy leveldiagram of 1 and the workfunctions of electrodes
82 H-C Su
323 Non-doped Near-Infrared LECs Basedon Interferometric Spectral Tailoring
Near-infrared (NIR) LECs show potential in NIR light sources because of simpledevice architecture compatibility with large-area solution processes low powerconsumption and high device efficiency Nevertheless host-guest NIR LECsgenerally exhibit enormously increased residual host emission when enhancing NIRlight output by raising bias [16] More NIR power was only harvested at the cost ofspectral purity To enhance light output from NIR LECs without losing spectralpurity a new way to obtain NIR emission from non-doped deep-red-emitting LECsby tuning active layer thickness to adjust microcavity effect has been reported [17]NIR emission from non-doped deep-red-emitting LECs was achieved by tuningemissive layer thickness to move the constructive interference peak to the NIRspectrum NIR emission coming from enhancement of microcavity effect wasinsensitive to bias Hence without sacrificing spectral purity 20X NIR output wasachieved when compared to reported value from host-guest NIR LECs [16]
NIR emission from non-doped deep-red-emitting LECs was achieved byemploying the deep-red-emitting 3 as the emissive material Complex 3 was [Ru(dtb-bpy)3][PF6]2 (where dtb-bpy is 44prime-ditertbutyl-22prime-bipyridine) [18] Thethickness of emissive layer was 605 nm As shown on Fig 36 LECs based onthick emissive layer finally showed a main NIR emission and a weaker red emis-sion However the EL spectrum of relatively thinner device (200 nm) showed astable peak at ca 660 nm Spectral shift in constructive interference because ofrecombination zone moving showed little effect on spectral changing Spacingbetween two constructive interference peaks in thinner reference LECs was muchlarger Constructive interference may take place outside the emission spectrum of 3rendering stable EL It showed that the output EL emission from non-dopeddeep-red-emitting LECs can be modified by tuning the emissive layer thickness toadjust microcavity effect
500 600 700 800 90000
02
04
06
08
10t (min)
15427596
EL In
tens
ity (a
u)
Wavelength (nm)
Fig 36 Temporal ELspectra of LECs with thickemissive layer (605 nm) of 3
3 Optical Engineering of Light-Emitting Electrochemical Cells hellip 83
To clarify the mechanism of the temporal EL from thicker LECs the proposedtechnique shown in Sect 311 was employed to probe moving of recombinationzone Time-dependent recombination zone for thick LECs based on 3 under 3 V isshown in Fig 37 The ruthenium(II) complexes with bipyridyl ligands werereported to exhibit higher electron mobility than hole mobility [19] Hence whenthe doped layers have not yet well established to facilitate carrier injection theinitial recombination zone was nearer anode because of imbalanced hole andelectron carrier mobilities Subsequent recombination zone moving was rationalizedby the energy levels depicted in Fig 38 Since injection barrier for hole (042 eV)was lower than that for electron (103 eV) the necessary number of ions near anodeto reach ohmic contact for hole was lesser than that for electron at the cathodeEnhancing rate for hole injection was thus higher than that for electron injectionleading to recombination zone shifting from anode toward the central active layer(Fig 37) After the doped layers were well established carrier injection efficiencyreached a steady state and the recombination zone froze It further revealed thattemporal EL was attributed to modified microcavity effect due to recombinationzone moving
Voltage dependent EL spectrum was the main disadvantage of NIR host-guestLECs As bias voltage enhanced to obtain higher light output residual host emis-sion was obvious deteriorating the spectral purity Low NIR power lt3 lWcm2
was achieved to avoid obvious host emission [16] Nevertheless the NIR ELspectra coming from spectral tailoring by tuning emissive layer thickness wasindependent on bias voltage The steady-state EL spectrum remained almostunchanged when the bias voltage increased from 3 to 35 V (Fig 39) This indi-cated that higher bias voltage speeded up device turn on but the steady-state
ITO
605 nm
456 nm
570 nm
PEDO
TPS
S
Ag
Ini al15 min
Final 96 min
Fig 37 Time-dependentrecombination zone in thickLECs (605 nm) containing 3at 3 V
Fig 38 Work functions ofelectrode metals and energylevels of complex 3
84 H-C Su
recombination zone position only changed slightly In comparison to host-guestNIR LECs [16] 20X NIR light power was harvested in thick LECs based on 3and spectral purity was not deteriorated significantly To reach higher and purerNIR output power modifying the EL spectra of non-doped deep-red-emitting LECsby altering microcavity effect was a feasible way
33 Outcoupling Extraction Technologies to EnhanceDevice Efficiencies of LECs
331 Enhancing Light Extraction by Employing MicrolensArray
As mentioned in Sect 312 much EL is trapped in the glass substrate because oftotal internal reflection To increase light extraction efficiency roughing theglassair interface to destroy total internal reflection is a possible way Edman et alreported enhancing device efficiency of LECs by employing light outcoupling films[20] Light outcoupling films were composed of hexagonal arrays of microlenses ona poly(methyl methacrylate) (PMMA) substrate With laminated light outcouplingfilms on the output side of the LEC and a highly reflective metallic film on the backside the outcoupled EL from LECs was enhanced by 60 In addition lightoutcoupling films randomized the output EL pattern and thus spatially inhomo-geneous EL in large-area LECs can be reduced This work demonstrated greatpotential of improving device efficiency of LECs by enhancing light outcouplingefficiency
500 600 700 800 90000
02
04
06
08
10
30 V 35 V
EL In
tens
ity (a
u)
Wavelength (nm)
Fig 39 Stabilized ELspectra of thick LECs(605 nm) based on 3 under 3and 35 V
3 Optical Engineering of Light-Emitting Electrochemical Cells hellip 85
332 Recycling the Trapped EL by Employing Red ColorConversion Layers
In addition to the light confined in glass substrate large portion of EL emission istrapped or waveguided in the ITO layer due to its high refractive index Recyclingthe EL in waveguided and substrate mode is a possible way to enhance deviceefficiency Su et al proposed LECs containing embedded red color conversionlayers (CCLs) to realize this concept [21] A CCL doped with a red-emitting dyewas placed under the ITO layer to absorb the blue emission confined in the glasssubstrate and in the ITO layer Then the absorbed blue EL turned into red PL whichcan be partially transferred to the external mode Since some trapped EL wastransferred into the external mode device efficiency can significantly be enhanced
The blue-emitting LEC with an embedded red CCL is shown in Fig 310Complex 1 was used in the emissive layer of the LEC The red CCL consisted of atransparent photoresist (TPR) film containing a red-emitting dye 4-(dicyano-methylene)-2-t-butyl-6-(1177-tetramethyljulolidyl-9-enyl)-4H-pyran (4) Thewaveguided blue EL in the ITO layer exhibited the evanescent field tail pene-trating into the organic and TPR layers Overlap between the evanescent field tailof the waveguided mode and the CCL (the shading area in Fig 310) facilitatedabsorbing of the blue EL in the ITO layer and transferring into the red PL whichcan be partially outcoupled for use In addition the blue emission trapped in thesubstrate mode can transmit into the CCL because of similar refractive indices ofglass and TPR Therefore the blue emission in the substrate mode can also bepartially turned into red PL in the external mode With recycled EL in waveguidedand substrate mode the measured maximal EQE (power efficiency) were up to125 (27 lmWminus1) These data were very high device efficiencies in white LECsand hence ensured that recycling trapped EL in waveguided and substrate mode iseffective in enhancing device efficiency of LECs
Ag
Complex 1
PEDOTPSSITO
CCL
Glass
Fig 310 Devicearchitecture of theblue-emitting LEC containingan embedded red CCL Profileof the optical field of thewaveguided mode is alsoshown for comparison
86 H-C Su
To further enhance device performance of LECs to realize practical applicationsTiO2 nanoparticles (NPs) were doped in red CCLs to induce scattering and thus toincrease EL outcoupling efficiency Red CCLs were composed of TPR films dopedwith 4 As shown in Fig 311 large NPs with diameter of 250 nm scattered andredirected the light propagating in red CCL and further increased the outcoupledlight while small NPs with diameter of 25 nm increased the effective refractiveindex of red CCL Similar refractive index of ITO layer and red CCL resulted in aless confined waveguide mode and more evanescent field tail in red CCL wasexpected leading to more scattering and transferring of the waveguide mode intothe external mode In addition the EL confined in glass substrate because of totalinternal reflection also propagated into the lower TPR layer and was then partiallyscattered into the external mode A higher concentration of DCJTB in the upperTPR layer enhanced turning the evanescent field tail of blue emission into red PLA lower concentration of DCJTB in the lower TPR layer avoided over absorption ofthe blue emission especially for the external mode and hence resulted in properblue and red intensity ratio Lower absorption of the external mode in the lower redCCL also improved device efficiency With enhanced outcoupling of trapped EL inwaveguided and substrate mode peak EQE (power efficiency) up to 20(40 lmW) was obtained Device efficiency was almost doubled by employing thistechnique These efficiencies were record values for white LECs
Temporal evolution in EL spectrum because of modified microcavity effectresulted from recombination zone moving was observed in white LECs containingred CCLs without scattering NPs [21] Microcavity effect commonly inducedangle-dependent EL spectrum from OLEDs [22] Scattering NPs in CCLs lead tomixed emission from all directions and spatially redistributed EL Therefore ratherstable EL spectra even when recombination zone was moving can be obtainedThe CIE coordinate migration (Dx Dy) of these LECs were only(plusmn 0002 plusmn 0005) Such stable white EL is essential in solid-state lighting
Glass
Upper CCL
ITO
Organic layer
Lower CCL
Fig 311 Device structure ofthe blue-emitting LECscontaining red CCLs dopedwith TiO2 NPs Schematicdiagram of the optical field ofthe waveguide mode is alsoshown for comparison
3 Optical Engineering of Light-Emitting Electrochemical Cells hellip 87
333 Recycling the Trapped EL by EmployingWaveguide Coupling
As mentioned in Sect 332 much EL was confined in ITO layer and thusextracting the trapped EL in ITO layer is a possible approach to enhance deviceefficiency of LECs However for LECs with CCLs only the emission in theevanescent field tail was extracted Most emission was still waveguided in the ITOlayer To extract the waveguided EL inside the ITO layer waveguide coupling wasemployed in white LECs based on 1 doped with 02 wt [Ir(ppy)2(biq)][PF6] (5)(where ppy is 2-phenylpyridine and biq is 22prime-biquinoline) [23] As shown inFig 312 two TPR layers containing TiO2 NPs were placed under ITO layer In theupper TPR layer the concentration of 25 nm TiO2 NPs was adjusted to modify theeffective refractive index and thus efficient waveguide coupling between ITO layerand the lower TPR layer was realized Because the lower TPR layer was doped with250 nm TiO2 NPs emission coupled from ITO layer was scattered and partiallytransferred into the external mode In addition the emission confined in glasssubstrate also propagated into the lower TPR layer and was partially recycled intothe external mode With extracted EL in waveguide and substrate mode deviceefficiency of LECs can be significantly improved
Directional couplers have been commonly used in integrated optics to spiltoptical intensity in waveguides Overlap of the evanescent field tails of the twowaveguide modes in the spacing layer (the shaded area in Fig 312) lead to opticalcoupling between the two waveguides [24] A directional coupler commonlyconsisted of two high-index waveguides separated by a thin low-index interlayer(the inset of Fig 313) Waveguide A and B in Fig 313 corresponded to the top
Spacing layer
Ag
Emissive layer
PEDOTPSS
ITO (Top waveguide)
Glass
Bo om waveguide
Fig 312 Device structure ofLECs employing waveguidecoupling Optical fieldintensity distributions of thewaveguide modes in ITOlayer and the lower TPR layerare also shown forcomparison
88 H-C Su
and bottom waveguides in Fig 312 respectively Optical intensity evolutionversus waveguide length in a directional coupler is shown in Fig 313 The opticalintensity transferred back and forth between the two waveguides It was physicallymodeled by the coupled mode theory [24] The dashed envelope line representedthe waveguide absorption Based on a similar waveguide architecture waveguidedlight in ITO layer can thus be coupled to the bottom waveguide layer Then the lightwas scattered and redirected into the external mode in consequence Effectiverefractive index of the spacing layer should be adjusted to increase the evanescentfield tail overlap of the two waveguide modes which enhanced waveguide cou-pling It was achieved by adjusting the concentrations of 25 nm TiO2 NPs in thespacing layer Refractive indices versus wavelength for ITO layer and spacinglayers doped with 25 nm TiO2 NPs in different concentrations are shown inFig 314 Higher concentration of 25 nm TiO2 NPs lead to higher effectiverefractive index of the spacing layer Lesser discrepancy in refractive index betweenITO layer and the spacing layer reduced the confinement in the optical field in ITOlayer and thus more evanescent field tail in the spacing layer can be expectedLarger evanescent field tails resulted in enhanced overlap between the twowaveguide modes rendering higher waveguide coupling efficiency With dopingconcentration of 25 nm TiO2 NPs of 8 peak EQE (power efficiency) up to 19(34 lmWminus1) was obtained It revealed that high device efficiencies of white LECswere achieved via waveguide coupling
Further increasing the refractive index of the spacing layer cannot improvedevice efficiency It resulted from that ITO layer cannot support the waveguidemode because of insufficient refractive index difference between ITO and spacinglayer (waveguide cutoff) The EL emission would be trapped in the ITO layerspacing layer and bottom waveguide layer Because waveguide coupling stoppedunder cutoff condition only EL in the bottom waveguide layer was recycled intothe external mode leading to deteriorated device efficiency The cutoff conditionie the minimal required refractive index difference between ITO and spacing layerto support the fundamental (lowest) mode in the ITO waveguide is shown in thefollowing [24]
Opt
ical
Inte
nsity
(au
)
z
1
0Waveguide Length
Waveguide AWaveguide B
Waveguide A
Waveguide B
Fig 313 Optical intensityevolution versus waveguidelength for a directionalcoupler Inset schematicdevice structure of adirectional coupler
3 Optical Engineering of Light-Emitting Electrochemical Cells hellip 89
DnITOSpacing layer frac14 k2016ethnITO thorn nSpacing layerTHORNd2
eth32THORN
where k0 is the vacuum light wavelength and d is the waveguide (ITO) thicknessWhen k0 = 490 nm was used for consideration the minimal DnITOSpacing layer forthe spacing layer with 16 25 nm-TiO2 NPs (n = 188) was calculated to be 026However the actual value of DnITOSpacing layer at 490 nm was only 018 which wasbelow the cutoff value For k0 gt490 nm the actual DnITOSpacing layer was evensmaller (Fig 314) and thus almost whole EL was under cutoff condition Becausewaveguide coupling stopped some emission was still confined in ITO layer andspacing layer resulting in deteriorated device efficiency It indicated that therefractive index of the spacing layer have to be carefully adjusted to ensurewaveguiding of ITO layer and thus to optimize the outcoupling efficiency
34 Conclusion and Outlook
Tremendous progresses have been made in device performance of LECs bydeveloping novel emissive materials and by improving carrier balance To furtherenhance device efficiency and to adjust optical properties of LECs optical engi-neering was a feasible approach to extract more confined EL in layered devicestructure In this chapter several reported optical engineering techniques have beenreviewed The microcavity effect can be employed to tailor the EL spectrum ofLECs due to altered interference effect in different device optical structuresReduced bandwidth of the sky-blue EL resulted from destructive interference atgreen can be achieved and hence purer white EL can be obtained without usingdeep-blue-emitting materials In addition recombination zone moving in LECsleads to constructive interference at red When combined with blue-green EL fromLECs white emission was realized in non-doped LECs based on a single emissivematerial Based on a similar concept non-doped NIR LECs containing a
500 600 700
16
18
20
22
24
Ref
ract
ive
Inde
x
Wavelength (nm)
ITO layer Spacing layer (4 25-nm TiO2) Spacing layer (8 25-nm TiO2) Spacing layer (16 25-nm TiO2)
Fig 314 Refractive indicesversus wavelength for ITOlayer and spacing layers with25 nm TiO2 NPs of 4 8 and16
90 H-C Su
red-emitting complex were reported to exhibit 20-fold enhancement in NIR lightoutput as compared to conventional host-guest NIR LECs To further improve lightoutcoupling efficiency microlens array was utilized on the output end of glasssubstrate in LECs and 16-fold enhancement in outcoupled EL was achievedRecycling the trapped EL by employing red CCLs placed under ITO layer would beanother effective approach The blue emission confined in the substrate mode and inthe waveguide mode can be transferred into red PL in the CCL and was partiallyredirected into the external mode With large NPs (250 nm) to scatter the emissionin the CCL peak EQE (power efficiency) of white LECs reached 20 (40 lmWminus1)Such results were almost doubled values when compared to white LECs withoutscattering red CCLs Waveguide coupling would also be an effective approach torecycle trapped light in the ITO layer Two TPR layers were placed under ITOlayer By tuning the effective refractive index of the upper TPR layer efficientwaveguide coupling between ITO layer and the lower TPR layer was realizedBecause the lower TPR layer was doped with scattering particles emission coupledfrom ITO layer was scattered and transferred into the external mode Doubleddevice efficiency can also be achieved by utilizing waveguide coupling
In spite of enormous enhancement in device performance achieved by usingoptical engineering techniques mentioned above further improvement will still berequired to meet practical applications Fabricating protrusion structures on andorbelow ITO layer would destroy waveguiding of light at the ITO layer renderingincreased light outcoupling efficiency Incorporating metal NPs in LECs would beanother way to enhance EL efficiency of LECs Emission of excitons can be sig-nificantly enhanced by localized surface plasmon resonance on metal NPs Inaddition overall consideration of interference effect in optical structure of LECseg optimizing the thickness of ITO poly(34-ethylenedioxythiophene)poly(styrene sulfonate) (PEDOTPSS) and active layer would be beneficial in maxi-mizing light outcoupling efficiency With these feasible techniques furtherimprovements in device efficiency will be expected
Acknowledgements The author acknowledges the financial support from Ministry of Scienceand Technology (MOST 105-2221-E-009-097-MY2)
References
1 X Liu D Poitras Y Tao C Py J Vac Sci Technol 22 764 (2004)2 TW Wang HC Su Org Electron 14 2269 (2013)3 A Chutinan K Ishihara T Asano M Fujita S Noda Org Electron 6 3 (2005)4 HC Su HF Chen FC Fang CC Liu CC Wu KT Wong YH Liu SM Peng J Am
Chem Soc 130 3413 (2008)5 L He J Qiao L Duan G Dong D Zhang L Wang Y Qiu Adv Funct Mater 19 2950
(2009)6 L He L Duan J Qiao G Dong L Wang Y Qui Chem Mater 22 3535 (2010)7 HC Su HF Chen YC Shen CT Liao KT Wong J Mater Chem 21 9653 (2011)
3 Optical Engineering of Light-Emitting Electrochemical Cells hellip 91
8 HC Su HF Chen PH Chen SW Lin CT Liao KT Wong J Mater Chem 22 22998(2012)
9 AB Tamayo S Garon T Sajoto PI Djurovich IM Tsyba R Bau ME ThompsonInorg Chem 44 8723 (2005)
10 RA Velapoldi HH Toslashnnesen J Fluoresc 14 465 (2004)11 GR Lin HF Chen HC Shih JH Hsu Y Chang CH Chiu CY Cheng YS Yeh HC
Su KT Wong Phys Chem Chem Phys 17 6956 (2015)12 CT Liao HF Chen HC Su KT Wong J Mater Chem 21 17855 (2011)13 CT Liao HF Chen HC Su KT Wong Phys Chem Chem Phys 14 9774 (2012)14 JS Lu JC Kuo HC Su Org Electron 14 3379 (2013)15 CL Lee CY Cheng HC Su Org Electron 15 711 (2014)16 CC Ho HF Chen YC Ho CT Liao HC Su KT Wong Phys Chem Chem Phys 13
17729 (2011)17 JH Hsu HC Su Phys Chem Chem Phys 18 5034 (2016)18 S Bernhard JA Barron PL Houston JL Ruglovksy X Gao GG Malliaras J Am
Chem Soc 124 13624 (2002)19 WK Chan PK Ng X Gong S Hou Appl Phys Lett 75 3920 (1999)20 N Kaihovirta C Larsen L Edman ACS Appl Mater Interfaces 6 2940 (2014)21 JS Lu HF Chen JC Kuo R Sun CY Cheng YS Yeh HC Su KT Wong J Mater
Chem C 3 2802 (2015)22 CC Liu SH Liu KC Tien MH Hsu HW Chang CK Chang CJ Yang CC Wu
Appl Phys Lett 94 103302 (2009)23 CY Cheng CW Wang JR Cheng HF Chen YS Yeh HC Su CH Chang KT
Wong J Mater Chem C 3 5665 (2015)24 RG Hunsperger Integrated Optics Theory and Technology (Springer New York 2009)
92 H-C Su
Chapter 4The Use of Additives in Ionic TransitionMetal Complex Light-EmittingElectrochemical Cells
Lyndon D Bastatas and Jason D Slinker
Abstract In this chapter we will describe the use of additives within films of ionictransition metal complexes (iTMC) applied in light-emitting electrochemical cells(LEC) Here iTMCs generally act as charge transporters for electrons and holeselectrolytes for ion motion and emissive materials The additives primarily fall intothree categories polymer additives small molecule hostguest systems andsaltelectrolyte additives Each class has emerged to optimize specific elements ofthe device operation
Keywords Ionic transition metal complex Lithium salts Host guest systems Ionic liquids Electrical double layer
41 Polymer Additives to Decrease Self-quenchingfor Improved Efficiency
In light-emitting electrochemical cells (LEC) electrons and holes injected from thecontacts traverse through the bulk and recombine to form excitonsmdashie boundelectron-hole pairs Ideally these excitons will decay radiatively and a chief benefitof LECs based on ionic transition metal complexes (iTMC) is their high luminanceefficiency which can approach 100 (see Chaps 7 8 and 11 for details about thistype of compounds) [1] One fundamental issue is the nonradiative self-quenching ofexcitons due to the presence of excimers andor exciplexes in solid state [2 3] Thatis when two or more excited-state molecules are in close proximity to each otherluminescence can be lost as the energy is dissipated between the molecules them-selves through a variety of nonradiative processes This is an inherent problem inpristine iTMC devices where the metal complexes function as mixed conductorsmdashie facilitating transport of electrons and holes as well as serving as an electrolyte
LD Bastatas JD Slinker (amp)The University of Texas at Dallas 800 W Campbell Rd RichardsonTX 75080 USAe-mail slinkerutdallasedu
copy Springer International Publishing AG 2017RD Costa (ed) Light-Emitting Electrochemical CellsDOI 101007978-3-319-58613-7_4
93
for counterions in the film [4 5] Compounding this fact the operational mechanismof these materials dictates that recombination of electrons and holes occurs within athin region between the electrodes and thus excitons are confined to a small space[6 7] These combined effects can lead to electroluminescence efficiencies in devicesthat are lower than anticipated from the photoluminescence quantum yields () ofthe materials found in solution
To counteract these effects efforts have been focused on spacing the metalcomplexes apart in various ways Intrinsically this has been accomplished throughthe addition of bulky ligands on the metal complexes themselves [8ndash10] Suchapproaches have led to some of the most efficient LECs produced to date Howeverthere is a limit to this approach as very large ligands can decrease both the elec-tronic and ionic conductivity of the films [11] Alternatively iTMC spacing can beincreased through additives and early efforts in iTMC films utilized inert polymersto accomplish this Prior photoluminescence studies had shown that inert polymerssuch as poly (methyl methacrylate) (PMMA) can enhance the of solid state filmsof ITMCs [12] This idea was then incorporated into some of the first iTMC deviceswith important implications for device efficiency and lifetime
411 Layer-by-Layer Techniques
To control complexndashcomplex spacing with inert polymer films the Rubner groupfirst employed layer-by-layer polymer assembly techniques to controllably depositthe electroluminescent layer of ruthenium(II) iTMC-based LECs [13ndash15] Thisinvolved the alternating deposition of acidic basic or neutral layers (Fig 41)which was accomplished either by hand or by automated dipping machines Suchassembly approaches are similar to LangmuirndashBlodgett film deposition and theapproach enabled precise control over the composition and thickness of thematerials In the first embodiments of this approach ruthenium(II) complexes wereintroduced into devices as pendant groups on polyester polymers and electricallyinert poly(acrylic acid) (PAA) layers were incorporated By changing the pH of theruthenium(II) polyester and PAA solutions during the deposition process both theoverall thickness of the devices and the relative ratio of the ruthenium(II) unit to thePAA spacer was varied The variation of the composition allowed the systematicmodification of the site-to-site distance of the ruthenium(II) complexes whichinfluences self-quenching effects and electronic mobility in the devices By opti-mizing thickness and composition external quantum efficiencies (EQE)mdashie theemitted photons per injected electron please refer to Chap 1 for more detailsmdashashigh as 32 were realized [15] a breakthrough at the time that remains highlyrespectable for ruthenium(II) iTMCs By simply changing the thickness of thesefilms at the optimal composition the EQE varied from 02 at 90 nm to 2 at200 nm active layer thickness The turn-on voltage was found to be a function ofthe number of polyesterPAA bilayers present with 3 and 5 V turn-on voltages for15 and 35 bilayer devices respectively This was attributed to the high bulk
94 LD Bastatas and JD Slinker
resistance of the multilayer films It was shown that the order of voltage sweepingwhether from forward bias to reverse or reverse bias to forward influenced thesymmetry of the current versus voltage and radiance versus voltage characteristicsFurthermore by changing the composition of the multilayers devices that exhibitedlight emission preferentially in either the forward or reverse bias operation werefabricated These first efforts focused primarily on polymerizable ruthenium(II)compounds but the Rubner group subsequently showed that a more simplifiedapproach could be followed
412 Blended Inert Polymers
iTMCs may be directly mixed with electrically inert polymers to achieve thebenefits afforded by a decrease of self-quenching Rudmann and Rubner firstexplored this idea with ruthenium(II) complexes mixed into inert polymers such aspoly(methylmethacrylate) (PMMA) poly(carbonate) (PC) and poly(styrene) (PS)discovering great improvements in lifetime [16] Inert polymers were introducedinto devices by combining stock solutions of the ruthenium(II) complex and theinert polymer and spin coating So this approach maintains the ease of solutionprocessing Polymer blending when coupled with pulsed voltage driving increasedthe lifetimes of the LECs considerably As seen in Fig 42 the extrapolated life-times of single layer devices were extended for some of the polymer blends Thedevice half-life the time to decay to half of the radiant flux maximummdashsee Chap 1for more details reached an excess of 1000 h for polycarbonate blends Likewisethe EQE was improved to as high as 25 for polymer blended devices postulatedto possibly come from a reduction of leakage current with the incorporation of theinert polymer
Additional efforts by Rudmann et al with polymer blending led to even higherefficiencies [17] Ruthenium(II) complexes were modified with ligands that pro-moted additional spacing between complexes as shown in Fig 43 This led toEQEs reaching as high as 55 the best among iTMC devices to that time
Fig 41 Structure of anemissive polymer use to makelayer-by-layer polymerblended LECs
4 The Use of Additives in Ionic Transition Metal Complex hellip 95
(Fig 44) Furthermore through the use of small counterions coupled with thecomplex turn-on times of less than one second could be achieved However thesedevices showed relatively low luminance typically 10ndash30 candelas per metresquared (cdm2) below the benchmarks typically desired for display (gt100 cdm2)and lighting applications (gt3000 cdm2)
Later the group of Bolink utilized the polymer blending strategy to achievedevices of high brightness and relatively long-lasting emission In particular a tris(47-diphyenyl-110-phenanthroline) ruthenium(II) complex was blended withPMMA to yield a device with appreciable luminance of nearly 400 cdm2 at apower efficiency of 19 lumens per watt (lmW) and a lifetime of approximately100 h [18] The increased performance over more simplified ruthenium(II) com-plexes was attributed to the hydrophobic nature of the ligands which would tend to
Fig 42 Light emissionversus time of pristineruthenium(II) tris-bipyridineLECs and those blended withpolystyrene (PS)polycarbonate (PC) andpolymethylmethacrylate(PMMA) Solid lines showextrapolations to the half-lifeof the emission Reprintedfrom [16] with the permissionof AIP Publishing
Fig 43 Structures of thesmall molecule complexesand side profile illustration ofa blended light emittedelectrochemical cellReprinted with permissionfrom [17] Copyright 2016American Chemical Society
96 LD Bastatas and JD Slinker
suppress water-based degradation reactions that had been shown to plague thesecomplexes (see Chaps 7 and 8 for more details) [19ndash21]
In conclusion the polymer blending strategy is one way to circumventself-quenching in iTMCs-based LECs for improved EQE and lifetime Howeverthe increased resistance introduced by the inert polymer limits the overall brightnessof these devices and also negatively impacts the power efficiency Thus there isbenefit to retaining small molecules throughout the LEC active layer to maintainhigh conductivity and brightness
42 HostGuest LECs to Control Color and ImproveEfficiency
As noted above self-quenching of iTMC-based LECs is an inherent concern par-ticularly for pristine films of unblended iTMCs [22] Blending the emitter with apolymer decreases this quenching and improves efficiency but at a cost to the overalldevice resistance such that absolute luminance and power efficiency both suffer Thisis due to the fact that these emitters are multifunctional chromophores that alsotransport electrons and holes through hopping and diffusion processes and increasingthe spacing between these molecules greatly frustrates this transport Conductingpolymers can be used to circumvent this problem but this ultimately changesthe class of devices and falls out of the scope of this chapter (see Chaps 1 and 10 fordetails about this aspect) Also this approach often changes the operational mecha-nism from that of a LEC to a more conventional light-emitting diode Notably theunique operational mechanism of LECs has enabled their use in beneficial archi-tectures such as functional cascaded panels [23 24] and electroluminescent nanofi-bers [25 26]
Fig 44 Luminance andEQE versus time of a LECfrom a ruthenium(II)tris-bipyridine complexblended withpolymethylmethacrylate(PMMA) Reprinted withpermission from [17]Copyright 2016 AmericanChemical Society
4 The Use of Additives in Ionic Transition Metal Complex hellip 97
Alternatively to overcome the challenges of self-quenching without a cost toconductivity while maintaining a LEC operational mechanism two or moreiTMCs may be combined in a hostndashguest approach In the case of two materialsone may primarily function for charge transport and the other typically of lowerbandgap is utilized as an emitter Alternatively separate conduction of electronsand holes may be facilitated by the distinctive materials given the details of theenergy levels This has been particularly utilized in blends of iTMC-based LECsThe fabrication of hostndashguest iTMC-based LECs is facilitated by the solutionprocessability of these materialsmdashie both components can be co-dissolved anddeposited from the same solvent Separating charge transport and emission pro-cesses in LECs circumvents the self-quenching that decreases the emission yield inorganics
The first instance of the hostndashguest approach in an iTMC device was fromHoisseini et al who doped an osmium(II) complex into a ruthenium(II) complex toachieve an improved efficiency of a technologically beneficial color [27] [Os(phen)3][PF6]2 where phen is 110-phenanthroline emits a deep red color butexhibits a low It was doped within a [Ru(bpy)3][PF6]2 matrix an orange emitterwith a higher In Fig 45 the photoluminescence spectra of pristine [Os(phen)3][PF6]2 and [Ru(bpy)3][PF6]2 films are shown along with those from small per-centages of [Os(phen)3][PF6]2 within [Ru(bpy)3][PF6]2 Energy transfer from theruthenium(II) complex to the osmium(II) complex was facilitated by the overlap ofthe ruthenium(II) complex emission with the osmium(II) complex absorption Thenormalized electroluminescence of various blends of these materials at 3 V oper-ation is also shown in Fig 45 and clearly the electroluminescence spectra are
Fig 45 Top Normalizedphotoluminescence spectra ofdoped and pristine films under470 nm excitation BottomNormalizedelectroluminescence spectrafrom devices made withdoped and pristine films at3 V (Inset) Bias dependenceof the emission spectrum forthe 1 doped film Reprintedwith permission from [27]Copyright 2016 AmericanChemical Society
98 LD Bastatas and JD Slinker
nearly identical to the photoluminescence spectra By doping 5 of guest to emitterweight (ww) nearly all of the emission originates with the osmium(II) complexguest In the inset it is shown that increasing the bias increases the emissioncomponent from the ruthenium(II) complex likely due to saturation of the osmium(II) dopant states from exponentially increased carrier injection Overall this workshowed that it is possible to tune emission color with small amounts of one iTMCwithin another
In Fig 46 the EQE of the ruthenium(II) complex host osmium(II) complexguest and guest 5 ww blended device is shown The efficiency of the osmium(II)complex device is essentially an order of magnitude below that of the ruthenium(II)complex device However blending the small amount of the guest in the hostproduced a device that is more efficient than either individual film at a respectableluminance of 220 cdm2 The improvement apparently arises from thedecreased self-quenching of the osmium(II) complex Thus in addition to colortuning this hostguest approach can improve device efficiency
The hostguest strategy was later applied to iridium(III) complexes withimpressive results Su and coworkers extended this by doping the green-emitting [Ir(dFppy)2(SB)][PF6] host complex with the orange-emitting [Ir(ppy)2(SB)][PF6]guest complex where ppy is 2-phenylpyridyl dFppy is a fluorine substituted ppyand SB is a spirobifluorene ligand [28] Previously these complexes were studiedfor their high efficiency and color tuning [10] The highest was obtained for filmswith guest concentrations of 25 ww at which concentration emission was foundto almost completely occur from the guest Before blending these complexesdevices based separately on the host or guest each showed a respectable EQE of71 [10] Upon blending an impressive peak EQE of 104 and power efficiencyof 368 lmW were obtained [28] As in the previous case the improvement in
Fig 46 Temporal evolutionof the EQE of pristineosmium(II) complexes andruthenium(II) complexesdevices as well as a guest 5ww blended device (Inset)dependence of peak efficiencyon dopant concentrationReprinted with permissionfrom [27] Copyright 2016American Chemical Society
4 The Use of Additives in Ionic Transition Metal Complex hellip 99
efficiency over either pristine film signified a decrease in self-quenching of emissionupon doping in a matrix
This group also utilized the doping strategy to achieve white light emissionbetween two iridium(III) complexes as seen in Fig 47 [29] In this instance thegroup combined a blue-green-emitting host with a red-emitting guestPhotoluminescence studies of blended films revealed that emission was roughlybalanced between host and guest at a guest concentration of 04wt Blends of theseemitters in single layer LECs showed that the relative emission could be tuned basedon the choice of voltage as previously noted by Hoisseini et al At low voltages(27 V and lower) emission primarily occurred from the red guest whereas at highervoltages emission became more balanced as presumably the guest emission statesbecame saturated Power efficiencies as high as 78 lmW were demonstrated at aluminance of approximately 10 cdm2 Color rendering indicesmdashCRIs see Chap 1for more detailsmdashof 80ndash81 were obtained over the range of operating voltages of27ndash33 V This group went on to use the host guest strategy to produce devices with
Fig 47 Top White lightelectroluminescence spectrafrom a hostguest LECoperating at various biasesBottom Current density andvoltage characteristics from ahostguest whitelight-emitting LEC Reprintedwith permission from [29]
100 LD Bastatas and JD Slinker
white electroluminescence having Commission Internationale de lrsquoEclairagemdashCIEsee Chap 1 for more detailsmdashcoordinates approaching (033 033) Overall apromising strategy the hostguest approach can produce high-efficient devices withcolor control It remains to be seen if high luminance (gt300 cdm2) at high efficiencycan be achieved with this approach
43 Ionic Additives for Improved Ion Redistributionand LEC Performance
431 Electric Double Layer Formation and ChargeInjection
As carefully explained in Chap 1 LEC operation is facilitated by the redistributionof ions in the film to form electrical double layers at the contacts [6 30ndash38] Inparticular ions rearrange into layers forming a dense (Helmholtz) and diffuse(Gouy-Chapman) layer at each electrode referred to as electric double layers(EDLs) This ionic rearrangement controls carrier injection and as a result thedevice response ranges from seconds to as long as days particularly for pristinesmall molecule LECs [4 5 39] It follows that controlling the details of ionicredistribution and double layer formation is key to idealize LEC performance
432 ElectrolytemdashSalt Combinations
Initial efforts to control ionic redistribution in iTMC-based LECs involved usingtraditional electrolyte systems in combination with salts to control performanceThe first approach was carried out by adding poly(ethylene oxide) (PEO) and a saltlithium triflate to ruthenium(II) complexes [40] similar to polymer-based LECsystems which utilize these components along with a conducting electrolumines-cent polymer This reduced the turn-on time (ton) of these LECs from 2 min to 30 sHowever these devices showed limited EQE well below 1 Potential reasons forthis limitation could be increased electrical resistance due to the spacing of thecomplexes as well as possible phase separation of the complexes and electrolyte
A second approach utilizing an explicit electrolyte additive was accomplished byblending a crown ether and a salt along with an emissive binuclear ruthenium(II)complex [41] In this case the pristine device did not show emission on the timescale of the experiment but emitted light with ton as small as 2 min with an optimalelectrolytesalt combination Again these devices showed limited external quantumefficiencies of 002 This approach though potentially offering better phasecompatibility than PEO between the emitter and electrolyte could result in plas-ticization of the films at high ether contents thereby limiting optimal conductivity
4 The Use of Additives in Ionic Transition Metal Complex hellip 101
of the film As seen at appropriate concentrations both of these methods elicitedfaster response from the devices However the EQE remained low though alter-native emitters could have improved this figure-of-merit
433 Ionic Liquids
Following efforts with electrolyte-salt blends it was shown that salts could bebeneficially added to iTMC-based LECs without the addition of an explicit elec-trolyte In this arrangement the ionic emitters also act as electrolytes to distributethe salt One of the widely used classes of materials to increase the conductivity ofthe active layer of LECs is ionic liquids (ILs) [42] ILs behave as ionically con-ductive molten salts at room temperature primarily because the component ions arestructurally uncoordinated They have high conductivity appropriate for chargetransport as well as a wide potential window for stability both favourable forimproving LECs
To understand how these materials improve LEC performance consider that theionic conductivity of fully dissociated ions is
rion frac14Xi
niliqi eth41THORN
where ni is the concentration microi the mobility and qi the charge of the ith ionicspecies Ionic liquids can improve the ionic conductivity by increasing the con-centration of mobile ions and by increasing the mobility of ions in the film In turnthis can facilitate the redistribution of ions that contribute to beneficial double layerformation as noted above Consequently ILs tend to decrease the turn-on timelower the applied voltage needed to achieve a given luminance and increase theefficiencies of LECs
4331 Ionic Liquids Shown to Decrease Turn-on Time
Parker et al first investigated the effect of ionic liquids on the radiant flux turn-ontime and lifetime of LECs based on an iridium(III) complex [43] In particular theyprepared devices based on [Ir(ppy)2(dtb-bpy)][PF6] (where ppy is 2-phenylpyridineand dtb-bpy is 44prime-di-tert-butyl-22prime-dipyridine) and the ionic liquid1-butyl-3-methylimidazolium hexafluorophosphate [BMIM][PF6] They found thatas the concentration of ionic liquid was increased from 0ndash025 volume fraction(vv) the time for the onset of light emission and the time to achieve maximumemission decreased significantly (Fig 48b) This followed the trend anticipatedsince the conductivity of the film increases with increasing IL concentration asgiven by Eq 41 Simultaneously the intensity of light emission increased onlyslightly with concentration indicating that these ions had little effect on improving
102 LD Bastatas and JD Slinker
the balance of injected carriers Unfortunately device lifetime was also marginallylowered with increasing ionic liquid concentration (Fig 48c) Nonetheless thisestablished a convenient generalizable way to improve the response time ofiTMC-based LECs that has been utilized in many works Similar effects weresubsequently observed by Slinker et al [44] and Lowry et al [45] in iridium(III)iTMC-based LECs of different colours
4332 High Ionic Conductivity Ionic Liquids Yield High PeakLuminance
Costa et al evaluated the effect of additional imidazolium-based ionic liquids withdiffering intrinsic ionic conductivities in LECs from [Ir(ppy)2bpy][PF6] [46]
Fig 48 Left Radiant flux versus time of Iridium(III) iTMC-based LECs with various amounts ofionic liquid Middle Time for onset of emission (ton) and time to radiant flux maximum (tmax)versus ionic liquid concentration in iridium(III) iTMC-based LECs Right Lifetime and totalemitted energy versus ionic liquid concentration in LECs with [Ir(ppy)2(dtb-bpy)][PF6] Reprintedwith permission from [43] Copyright 2016 American Chemical Society
4 The Use of Additives in Ionic Transition Metal Complex hellip 103
Specifically the ILs investigated were [BMIM][PF6] 1-ethyl-3-methylimidazoliumhexafluorophosphate [EMIM][PF6] and 1-hexyl-3-methylimidazoliumhexafluorophosphate [HMIM][PF6] The intrinsic conductivity of each IL and theresulting device properties are provided in Table 41 Devices containing ILresponded more than one order of magnitude faster than the device without ILMoreover studies also confirmed that the turn-on time of the devices decreased withthe concentration of the IL added In addition they observed that there is a correlationbetween the luminance of the devices and their intrinsic conductivity with the peakluminance of more conductive devices higher than less conductive devices
The effect of ILs on turn-on times and luminescence were also observed by Parket al studying similar ILs They compared ILs with different counterions namely[BMIM][PF6] and [BMIM][BF4] in LECs with 11 emitter-to-IL molar ratio [47]The conductivity of the luminous layer was 122 158 and 166 nanosiemens percentimeter (nScm) for the pristine film the film with [BMIM][PF6] and a [BMIM][BF4] film respectively They observed that the turn-on time of the devices wasdictated by the conductivity of the film The peak luminance of the devices alsoincreased with the addition of the ILs and the device containing [BMIM][BF4]showed the highest peak They also confirmed the trends of the effect of [EMIM][PF6] and [BMIM][PF6] as noted above by Costa et al with a different iridium(III)complex in a separate work [48] Recently Sun et al observed decreased currentsand marginal efficiency changes in IL-doped LECs based on iridium(III) complexbut these differences could arise from the high thickness of the devices (600 nm)and relatively large IL concentrations used (see Chap 5 for more details) [49]
Table 41 Conductivities of various ionic liquids and their effects on LECs based on iridium(III)complexes Reprinted with permission from [46]
IL Intrinsicconductivity(mScm)
iTMCIL
tond
(min)t12e
(h)Luminance(cdm2)
PowerefficiencylmW
Etotf
(J)
[EMIM+][PF6
minus]52a 41 24 81 497 46 67
11 29 41 615 49 06
[BMIM+][PF6
minus]15b 41 82 103 269 86 23
11 35 43 586 167 02
[HMIM+][PF6
minus]11c 41 132 134 153 56 11
11 22 55 302 141 01
without 10 690 668 219 61 66aValue at 29815 K from ref 23bValue at 29815 K from Ref 25cValue at 2951 K from Ref 24dTime to reach 100 cdm2
eTime to reach half of the maximum luminancefTotal emitted energy calculated by integrating the radiant flux curve from the application of currentto the time for the radiant flux to decay to 15 of maximum Measured from a 3 mm2 device
104 LD Bastatas and JD Slinker
4333 Ionic Liquids in LECs Under AC and Pulsed Operation
Slinker et al demonstrated that additives in iTMC-based LECs can enable deviceswith direct operation from a standard US outlet at 120 V and 60 Hz [50] such asone seen in Fig 49 To enable the millisecond response required by 60 Hz theyincorporated [BMIM][PF6] into a ruthenium(II) iTMC-based LEC to enhance theconductivity of the light-emitting layer To accommodate the large applied biasthey utilized a cascaded device geometry in which the cathode of one device servedas the anode of an adjacent device in series The rapid response afforded by theionic liquid supported sufficient ionic redistribution for light emission to consis-tently appear on forward and reverse sweeps of the applied field Tordera et alshowed that the use of ionic liquids with the application of pulsed current canproduce an almost instantaneous (lt1 s) turn-on with high stability [51] Thesedevices operated at 1000 Hz with a 30 duty cycle and initially emitting670 cdm2 showed extrapolated lifetimes longer than 4000 h Such devices areamong the longest lasting iTMC-based LECs reported to date
434 Lithium Salt Additives
iTMC-based LECs are typically made with cationic ionic transition metal com-plexes and associated negatively charged counterions Generally the transitionmetal complexes are bulky ions that have larger radii and higher atomic weightsthan the counterions The larger size of the complex ion affects the net ionicconductivity as well as their spatial distribution in response to electric fields Underan applied bias the complexes are considerably less mobile than the counterions
Fig 49 Ionic liquidsenabled LEC operationdirectly from a US outlet at120 V and 60 Hz Photocourtesy of Prof JasonSlinker
4 The Use of Additives in Ionic Transition Metal Complex hellip 105
and are generally considered stationary This usually results in LECs that exhibitlong turn-on times Furthermore due to imbalanced conductivity and packingdensity this size difference presumably leads to a lower accumulation of uncom-pensated cations near the cathode relative to anions near the anode Consequentlythe interfacial electric field at the cathode is lower than the field in the anode(Fig 410 left) The greater accumulation of counterions at the anode interfacefavours hole injection making it difficult to achieve balanced carrier injection in apristine iTMC LEC
One simple way to address the challenge above is incorporating a salt additivethat utilizes small cations [52ndash54] The ideal added cations should compensate thelow mobility of the bulky complex cations and redistribute efficiently to increasethe space charge at the cathode more closely matching the space charge at theanode as illustrated in Fig 410 Ionic additives may also minimize aggregation ofthe emitters that can cause self-quenching and decrease the luminescence of thedevice However for real devices there is a threshold concentration for the bene-ficial effects of salt addition within an LEC [54] Adding more salts above thisconcentration could potentially result in side effects that hinder optimal deviceperformance Like the ionic liquid approach above the salt-additive strategy doesnot utilize an added associated electrolyte but rather utilizes the inherent iontransporting properties of the ionic transition metal complexes in the film Thisleads to good phase compatibility and ease of processing
Fig 410 Illustration of ionic charge redistribution under an external electric field for pristine(left) and salt-additive (right) devices Upon application of bias the cations move towards thecathode and anions towards the anode Without the salt additive a denser accumulation of anionsat the anode leads to a higher interfacial field than at the cathode Adding salts with small-sizecations promotes denser packing of cations at the cathode which evens the packing of anions at theanode leading to balanced fields at the contacts Reprinted from [52] with the permission of AIPPublishing
106 LD Bastatas and JD Slinker
4341 Lithium Additives Improve ITMC-Based LECs
The effect of direct salt additives in iTMCs was tested by Shen et al in LECs from [Ir(ppy)2(bpy)][PF6] (Fig 411) [52] Anticipating that the packing of [Ir(ppy)2(bpy)]
+
(rIr = 126 Aring) at the cathode was less dense than the packing of [PF6]minus at the anode
various salt additives with different cationic sizes were investigated The cationsexplored were [NH4]
+ (rNH4 = 143 Aring) K+ (rK = 133 Aring) and Li+ (rLi = 076 Aring)with each using the same anion as the complex [PF6]
minus Simple single layer deviceswere prepared with films blended with small amounts (01ndash033 wt) of the saltadditives All of the salts improved the luminance maximum as seen in the Fig 412plot of luminance and efficacy versus LEC operational time For [NH4]
+ and K+
additives there was a fractional increase (20 and 30 respectively see upper panel)in the absolute luminance However the Li+ additive produced a substantial increaseof luminance reaching about 5000 cdm2 and efficacy each fourfold over thepristine device Luminance and efficacy consistently increased as cationic radius ofthe salt additive decreased Overall every efficiency metric was improved with saltaddition as shown in Table 42
The response timemdashinversely related to ton see Chap 1 for more detailsmdashwasalso improved with addition of lithium salt while ammonium and potassium saltsdid not appreciably change ton The cationic radii of [NH4]
+ and K+ are about twice
Fig 411 Lithium additives for the enhancement of luminance of single layer iTMC LECs Thestructures of [Ir(ppy)2(bpy)][PF6] (top) and [Ir(DiPhPy)2(bpy)][PF6] (bottom) are shown materialsshown to yield high luminance along with the resulting light-emitting devices Illustrationcourtesy of Lyndon Bastatas Dr Kristin Suhr and Prof Jason Slinker
4 The Use of Additives in Ionic Transition Metal Complex hellip 107
as large as Li+ and likely affects the mobility of the ions hindering effectiveredistribution As for Li+ while the pristine device required 49 h to achievemaximum luminance this value was reduced by tenfold upon adding Li[PF6] andheat processing reduced ton to a mere 10 s
Fig 412 Luminance andefficacy of [Ir(ppy)2(bpy)][PF6] LECs with blended saltsof various cations driven withconstant 0050 Acm2 currentdensity Notably the effect ofthe lithium salt in the deviceperformance stands outamong the salts consideredReprinted from [52] with thepermission of AIP Publishing
Table 42 Performance metrics of iridium(III) iTMC-based iTMC LECs with blends of varioussalts at 01wt unless otherwise noted Reprinted from [52] with the permission of AIPPublishing
Additive tona
(h)t12b
(h)Etotc
(J)Utotd
(Jmm2)Currentefficiency(cdA)
Powerefficiency(LmW)
Quantumefficiency(phel )
Luminancemaximum(cdm2)
None 49 167 345e 115e 24 19 077 1180
[NH4][PF6]
77 199 321e 107e 28 22 092 1410
K[PF6] 63 295 554e 185e 31 23 101 1560
Li[PF6] 46 37 182 61 99 58 321 4950
Li[PF6]f 00028 137 338 113 60 38 197 3030
aTime from application of current to maximum radiant fluxbTime from ton to 12 of radiant flux maximumcTotal emitted energy calculated by integrating the radiant flux curve from the application of current tothe time for the radiant flux to decay to 15 of maximum Measured from a 3 mm2 devicedTotal emitted energy density calculated by dividing Etotby the device active areaeExtrapolated values assuming first order exponential decay of the radiant fluxfDevice cast from a solution that was heated at 65 degC for 10 min 033wt salt
108 LD Bastatas and JD Slinker
This simple salt addition approach to improving iTMC-based LECs has manybenefits Only very small amounts generally 01ndash05wt of the salts are needed toimprove the device performance Hence this is a low-cost approach for large-scaledevice fabrication The approach maintains solution processing as the salts can beco-dissolved in the common polar solvents used with iTMCs The ionic nature ofboth the emitter and the additive promotes good phase compatibility The techniqueis also generalizable to many iTMCs as will be discussed below
If there is any drawback to Li[PF6] addition it appears that there is a trade-offbetween the benefits of fast response and high luminance versus the drawback of ashorter lifetime However the lifetime measurement must be carefully interpretedbecause the devices at various salt concentrations can operate at significantly dif-ferent luminance levels complicating the interpretation of the lifetimesmdashseeChap 1 for more details In such cases it is helpful to utilize the lifetimesextrapolated at a consistent luminance across devices Following industry standardsthe extrapolated lifetime is given by the acceleration equation
T2 frac14 T1L1L2
AF
eth42THORN
where T2 is the extrapolated lifetime at a maximum luminance of L2 T1 is themeasured lifetime at a maximum luminance of L1 and AF is a dimensionlessexponential acceleration factor generally taken to be 15ndash16 [53] EmployingEq 42 at 1000 cdm2 luminance level and a modest AF = 15 the extrapolatedlifetimes for all devices added with salt is higher than the salt-free device by morethan 100 h Furthermore extrapolated lifetimes at 100 cdm2 a common bench-mark for displays reaches over 10000 h for the Li+ additive devices studied Alsoit can be argued that the high luminance from salt additives did not compromise theemitter stability as the devices maintain almost the same energy output
Optimal luminance and response dynamics can be hindered by incompletedissociation of ions in solution Similarly modelling study suggests that a highbinding energy plays a role in the slow response of polymer-based LECs [55]Contributing to this effect is also the possible short-range crystallization [56] andaggregates in the film that discourage ionic distribution (see Chap 6 for moredetails) It was postulated that heating the solution before spin coating couldimprove performance by encouraging ionic dissociation and limiting crystallizationand aggregation For salt-free devices heating the pristine solution at 65 degC for tenminutes increases the luminance of the LEC device by threefold and reduces theturn-on time from 25 days to 25 h Applying this heating technique to a saltblended solution yielded devices with 10 s turn-on times and half-lives of over100 h (Fig 413) Thus the details of film processing with salt additives can bevery influential in controlling LEC performance
Lithium salt additives were found to enhance LEC performance from otheriTMCs In particular improved performance was also observed in [Ir(DiPhPy)2(bpy)][PF6] where DiPhPy is a 24-diphenylpyridine ligandmdashFig 411
4 The Use of Additives in Ionic Transition Metal Complex hellip 109
[53] Under constant current the devices with salt additive exhibited a peakluminance that is 50 higher than salt-free device reaching as high as 5500 cdm2
upon addition of salt The extrapolated lifetimes of devices at 1000 and 100 cdm2
luminance level were 120 plusmn 10 h and 3800 plusmn 400 h respectively The maximumEQE showed a modest increase from 345 for the pristine device to 482 withaddition of salt Similarly the efficacy was also improved by Li[PF6] from 72 to109 cdA Likewise a significant increase in power efficiency of 87 wasobserved Following this Li[PF6] was used to enhance a series of phenyl-cappediridium(III) complexes-based LECs [57] These works demonstrated that the ben-eficial effect of Li[PF6] additives was generalizable to other small molecule LECs
4342 Scanning Probe Study of Lithium Salt Additives
The effect of lithium salt additives on iTMC-based LECs has been experimentallyrevealed with scanning probe microscopy study on planar devices [38] Asexplained in Chaps 1 and 2 this approach is particularly advantageous for studyingmixed conductors where ionic redistribution leads to unique potential profiles anddrastic changes in local potentials with time [35] The confined geometries ofsandwich-structure devices (100 nm thick films between electrodes) restrict theuse of these direct probe techniques but the open geometry of planar devices withrelatively large (gt1 micron) interelectrode spacing works well The active film inthese devices should be meticulously patterned to capture the ion-blocking featureof electrodes [35 38] Common scanning techniques to probe the potential distri-bution in the active film of the device include electric force microscopy (EFM) andscanning Kelvin probe microscopy (SKPM)
Along these lines Lin et al performed SKPM to investigate the effect of the Li[PF6] ionic additive on the dynamics of planar devices with channel lengthsof 10 microm [38] In particular the impact of lithium salt additives on the doublelayer formation of these LECs was investigated in detail Devices were prebiased at
Fig 413 Comparison of thelight output of devices madefrom a warmed [Ir(ppy)2(bpy)][PF6] + 033wtLi[PF6] solution and from apristine film cast fromunheated solution Reprintedfrom [52] with the permissionof AIP Publishing
110 LD Bastatas and JD Slinker
50 V to provoke a significant ionic redistribution then tested by SKPM under 9 Vapplied bias to reveal differences in voltage profiles (Fig 414) For the pristinedevice the potential between the electrodes resembles a simple resistor Uponcloser inspection the potential proceeding from the cathode follows eminusjx depen-dence that extends throughout the bulk consistent with a diffuse layer of cationsThis demonstrated that the [Ir(ppy)2(bpy)]
+ do not accumulate with high density atthe cathode Alternatively devices with Li[PF6] showed a sharp linear drop at thecathode consistent with a dense accumulation of Li+ and reflecting more idealizeddouble layer formation Notably both devices showed similar dynamics at theanode as both had strictly [PF6]
minus anionsThe spatial electric field profiles for the same planar devices are shown in
Fig 415 Both devices show similar electric fields near the anode due to accu-mulation of [PF6]
minus At the cathode there is a much stronger peak electric field in theionic additive device due to the influence of densely packed Li+ Also indicative ofideal ionic redistribution the electric field is repressed in the bulk region (12 to16 lm) in the ionic additive device relative to the pristine device In ideal LECsions accumulate at the contacts until the electric field is effectively cancelled out by
Fig 414 Steady-statepotential distribution of planariTMC-based LECs at 9 Voperation SKPM data(voltage versus position) ofpristine (upper graph green)Au[Ir(ppy)2(bpy)][PF6]Audevices and devices withlithium salt additive (lowergraph red) Au[Ir(ppy)2(bpy)][PF6] + 05wtLi[PF6]Au under a 9 V biasDevices were also subjectedto a 50 V prebias for 30 minDashed lines represent theapproximate location of thegold electrode edges Theinset is the overlap range from7 to 21 lm positions on thegraphs Reprinted withpermission from [38]Copyright 2016 AmericanChemical Society
4 The Use of Additives in Ionic Transition Metal Complex hellip 111
the ionic space charge In real devices low mobility ions such as [Ir(ppy)2(bpy)]+
may persist in the bulk away from the contacts even with a substantial electric fieldin the bulk Addition of Li[PF6] contributes small cations that redistribute efficientlyto cancel out this electric field
It is informative to consider differences in the voltage profiles observed inruthenium(II) complex based devices which were extensively studied among thefirst iTMC-based LECs with those observed in LECs with iridium(III) complexeswith and without the presence of lithium salts We performed EFM measurementsof Au[Ru(bpy)3][PF6]2Au planar devices [35] However unlike the pristinedevice which shows a gradual potential drop from the cathode significant linearpotential drops at both electrodes were observed in pristine ruthenium(II) complexdevices This demonstrated that ion redistribution is more efficient and idealized indevices with ruthenium(II) complexes and clarifies why the turn-on times of thesepristine LECs are shorter than those from pristine iridium(III) complex devices Ionconduction in [Ru(bpy)3][PF6]2 devices benefit from two mobile counterions asopposed to one for [Ir(ppy)2(bpy)][PF6] and the ruthenium(II) complex is inaddition a smaller iTMC This supports the postulate that a difference in ion sizescan impact ion transport and packing densities of ions near the contacts resulting indifferences in interfacial fields that impact device efficiency It also shows thatadditives are most needed in low mobility films in LECs such as those from highlyefficient iridium(III) complexes Overall this work established that Li[PF6] addi-tives enhance double layer formation for improved iTMC-based LEC performance
4343 Optimal Lithium Salt Concentration
Modelling study suggests that a high mobile ion concentration is needed to opti-mize the device performance [58] Such device films have an enhanced ionic
Fig 415 Profile of electricfield in Au[Ir(ppy)2(bpy)][PF6]Au (green) and Au[Ir(ppy)2(bpy)][PF6] + 05 wtLiPF6Au (red) planar devicesapplied with 9 V Reprintedwith permission from [38]Copyright 2016 AmericanChemical Society
112 LD Bastatas and JD Slinker
conductivity of the active layer that results in an increased steady state currentdensity and recombination rate yielding enhanced charge injection and a narrowrecombination zone On the other hand devices with low concentrations of mobileions can exhibit injection limited current and extremely long turn-on times due toincomplete formation of one or both EDLs near the contacts Howeverdoping-induced quenching finite salt solubility and side reactions can occur whichcan set the threshold amount of salt that can be added to beneficially affect thedevices
The concentration dependence of Li[PF6] sandwich-structure LECs with iridium(III) complexes has been explicitly reported [54] Figure 416 shows the absoluteluminance of the devices with different weight fractions of Li[PF6] as well asrelative luminance and relative lifetimes as a ratio against the performance ofsalt-free device As shown for lower concentrations addition of increasingamounts of Li[PF6] promotes faster response kinetics higher peak luminance andcorresponding better maximum efficiency metrics than pristine devices Above athreshold concentration this trend reverses and performance suffers with increasingLi[PF6] Among the concentrations considered LECs with 05 wt Li[PF6]emerged as the optimal device composition At this concentration the devicesdisplayed almost three times the maximum brightness at only half the turn-on timerequired of pristine devices Although the absolute half-life of the devices with 05wt of Li[PF6] appears to be lower than the pristine device these are difficult todirectly compare since the devices were operating at different luminanceAlternatively for a more balanced comparison the extrapolated half-life at100 cdm2 is 35 times longer for the 05 wt lithium salt-additive device than the
Fig 416 The performance of iridium(III) iTMC-based LEC devices with various amounts of Li[PF6] under constant current driving (0050 Acm2) Left Luminance versus time for iridium(III)iTMC-based LECs with various Li[PF6] concentrations under constant current driving RightRelative maximum luminance and relative lifetime extrapolated at 100 cdm2 versus Li[PF6]concentration for LECs with iridium(III) complexes each normalized against a pristine LEC Errorbars represent the standard deviations over 4 experimental trials Reprinted with permission from[54] Copyright 2016 American Chemical Society
4 The Use of Additives in Ionic Transition Metal Complex hellip 113
pristine device (Fig 416 right) Surface probe measurements of planar devicessuggested that this enhancement is linked to details of double layer formationFurther measurements of sandwich-structure devices allowed for quantitative esti-mates of this effect
4344 Electrochemical Impedance Spectroscopy of LECswith Lithium Salt Additive
To understand the effect of different lithium salt concentrations electrochemicalimpedance spectroscopy was conducted at 0 V to follow ion-dominated chargetransport As explained in detail in Chap 1 electrochemical impedance spec-troscopy (EIS) enables the direct study of thin film sandwich-structure devices andextraction of key metrics of electronic and ionic motion EIS analysis showed thatthe conductance of the device is boosted by small additions of salt Initially con-ductance increases as more salt is added in small fraction peaking at the optimalconcentrationmdashie 05 wt Li[PF6] then decreases as more salt is added above thethreshold concentration This is corroborated by the opposite trend of resistancedisplayed in the plateau of the Bode plot shown in Fig 417 To further understandthis trend devices were modelled with an equivalent circuit that allows extractionof the thickness of EDL (Fig 417) Apparently there is a correlation of the deviceefficiency and initial packing of cations in the cathode A plot of the normalizedefficiency versus inverse double layer thickness shown in Fig 418 exhibits linearbehaviour with an R-squared correlation of 0946 To understand this correlationconsider that EDLs need to form to enhance electronic injection In particular tooptimize the device performance the potential barrier as influenced by the EDLthickness must be thin enough for easy tunnelling of injected electronic charges(Eq 41) to maintain steady charge balance between the contacts and efficient lightemission in the bulk of the device Thus it appears that with low Li[PF6] con-centration Li+ salt cations compensate the imbalance of space charge near thecathode in the pristine device to balance the effect of accumulated anions at theanode Alternatively at high Li[PF6] concentrations efficiency drops as thecathodic double layer broadens [38] This broadening could arise from aggregationof Li[PF6] which would plausibly lower ionic conductivity and limit cationaccumulation at the cathode Simulation studies showed that at high concentrationsin an ionic liquid environment Li[PF6] becomes more viscous and forms aggre-gates that decreases its conductivity and slows down ionic transport [59]Additionally high concentrations of Li[PF6] could lead to a lowered efficiency ofdissociation of the salt into mobile ions As pointed out earlier although addition ofsalts can improve the complex spacing and minimize self-quenching excessive
114 LD Bastatas and JD Slinker
Fig 417 Electrochemical impedance spectroscopy of thin film iridium(III) iTMC-based basedLECs with various fractions of lithium salt additives Left Equivalent circuit used for EIS fittingLcab is the inductance of the external cables Rext is the external resistance RE is the total electricalresistance of the active layer CPEGEO is a constant phase element from which is derived thegeometric capacitance Rion is the bulk resistance CEDLA and CEDLC are capacitors representing theelectrical double layers and REDLA and REDLC are the resistances of the double layers RightImpedance versus frequency data for iridium(III) iTMC-based LEC devices with variousconcentrations of Li[PF6] Solid lines are fits to the data based upon the equivalent circuit in part A(These are hard to see given the excellent agreement with the data) Reprinted with permissionfrom [54] Copyright 2016 American Chemical Society
Fig 418 Normalized EQEand inverse cathodic doublelayer thickness versus Li[PF6]concentration in iridium(III)iTMC-based LECs Reprintedwith permission from [54]Copyright 2016 AmericanChemical Society
4 The Use of Additives in Ionic Transition Metal Complex hellip 115
separation can compromise electronic charge hopping between complexes andnegatively impact ionic charge transport Thus an appropriate concentration ofadded ions must be used in order to observe beneficial effects
4345 Counterion Dependence of Lithium Salts
Most studies of lithium salt additives to date have utilized Li[PF6] since the [PF6]minus
counterion is commonly used to balance the positive charge of iTMCs RecentlyBandiello et al investigated LECs with lithium salts of various counterions andobserved distinct performance patterns from each [60] In addition to Li[PF6]perchlorate ([ClO4]
minus) tetrafluoroborate ([BF4]minus) and triflate ([CF3SO3]
minus) lithiumsalts were each studied and blended with iridium(III) complexes Interestingly theturn-on times from the LECs modified with these salts were inversely proportionalto the ionic conductivity of the salts found in solution This clearly demonstrates theimportant role that these salts play in ionic distribution and LEC turn-on timeLuminance and efficiency metrics were all improved approximately 30 by thelithium salts over the pristine device with the salt-enhanced luminance valuesreaching approximately 1000 cdm2 Interestingly three salts produced extensionsof lifetimes and two showed over tenfold enhancement namely Li[PF6] (1263 h)and Li[BF4] (1973 h extrapolated) This work represents further evidence of theability of lithium salts to produce bright long-lasting devices
44 Outlook
To date additives have played a key role in optimizing iTMC-based LEC perfor-mance The impact has been far reaching from improving key performance metricssuch as brightness efficiency and lifetime but also to enable interface with prac-tical technologies such as direct interface with ac line power [50] The future posesintriguing possibilities for additives in iTMC-based LECs As an example of apotential future application Fig 419 highlights the electroluminescence fromelectrospun fibres of microscale dimensions These fibres are based on a [Ru(bpy)3][PF6]2 emitter in a PMMA host similar to those described in a seminal publication[25] Such devices can be implemented for high resolution spectroscopy or inte-grated into lab-on-a-chip applications Clearly the landscape for these low-costdevices is broad and the focus for applications will sharpen as fundamentalunderstanding yields superior device performance
116 LD Bastatas and JD Slinker
Acknowledgements The authors would like to thank The University of Texas at Dallas for salarysupport for this effort The authors would also like to thank Dr Kristin Suhr and Prof JoseacuteMoran-Mirabal for assistance with the figures
References
1 H Yersin Top Curr Chem 241 1 (2004)2 E Margapoti V Shukla A Valore A Sharma C Dragonetti CC Kitts D Roberto M
Murgia R Ugo M Muccini J Phys Chem C 113 12517 (2009)3 JM Fernandez-Hernandez S Ladouceur Y Shen A Iordache X Wang L Donato S
Gallagher-Duval M de Anda Villa JD Slinker L De Cola E Zysman-Colman J MaterChem C 1 7440 (2013)
4 SB Meier D Tordera A Pertegaacutes C Roldaacuten-Carmona E Ortiacute HJ Bolink Mater Today17 217 (2014)
5 JD Slinker J Rivnay JS Moskowitz JB Parker S Bernhard HD Abruntildea GGMalliaras J Mater Chem 17 2976 (2007)
6 JC de Mello Phys Rev B 66 235210 (2002)7 J Gao J Dane Appl Phys Lett 84 2778 (2004)8 JD Slinker AA Gorodetsky MS Lowry J Wang S Parker R Rohl S Bernhard GG
Malliaras J Am Chem Soc 126 2763 (2004)
Fig 419 Electroluminescence from two electrospun nanofibres utilizing a [Ru(bpy)3][PF6]2emitter in a PMMA host on interdigitated electrodes with a one micron pitch Electroluminescenceis seen in the area between the electrodes Photo courtesy of Prof Jose Moran-Mirabal
4 The Use of Additives in Ionic Transition Metal Complex hellip 117
9 HJ Bolink E Coronado RD Costa E Ortiacute M Sessolo S Graber K Doyle MNeuburger CE Housecroft EC Constable Adv Mater 20 3910 (2008)
10 H-C Su F-C Fang T-Y Hwu H-H Hsieh H-F Chen G-H Lee S-M Peng K-TWong C-C Wu Adv Funct Mater 17 1019 (2007)
11 JA Barron S Bernhard PL Houston HD Abruna JL Ruglovsky GG MalliarasJ Phys Chem A 107 8130 (2003)
12 K Nagai N Takamiya M Kaneko J Photochem Photobiol A 84 271 (1994)13 JK Lee DS Yoo MF Rubner Chem Mater 9 1710 (1997)14 AP Wu JK Lee MF Rubner Thin Solid Films 329 663 (1998)15 AP Wu DS Yoo JK Lee MF Rubner J Am Chem Soc 121 4883 (1999)16 H Rudmann MF Rubner J Appl Phys 90 4338 (2001)17 H Rudmann S Shimada MF Rubner J Am Chem Soc 124 4918 (2002)18 HJ Bolink L Cappelli E Coronado M Graumltzel MK Nazeeruddin J Am Chem Soc
128 46 (2006)19 G Kalyuzhny M Buda J McNeill P Barbara AJ Bard J Am Chem Soc 125 6272
(2003)20 LJ Soltzberg JD Slinker S Flores-Torres DA Bernards GG Malliaras HD Abruntildea J-
S Kim RH Friend MD Kaplan V Goldberg J Am Chem Soc 128 7761ndash7764 (2006)21 JD Slinker J-S Kim S Flores-Torres JH Delcamp HD Abruntildea RH Friend GG
Malliaras J Mater Chem 17 76ndash81 (2007)22 S Bernhard JA Barron PL Houston HD Abruntildea JL Ruglovksy XC Gao GG
Malliaras J Am Chem Soc 124 13624 (2002)23 DA Bernards JD Slinker GG Malliaras S Flores-Torres HD Abruntildea Appl Phys Lett
84 4980 (2004)24 JD Slinker J Rivnay JA DeFranco DA Bernards AA Gorodetsky ST Parker M
P Cox R Rohl GG Malliaras S Flores-Torres HD Abruntildea J Appl Phys 99 074502(2006)
25 JM Moran-Mirabal JD Slinker JA DeFranco SS Verbridge GG Malliaras HGCraighead Nano Lett 7 458 (2007)
26 Z Zhang K Guo Y Li X Li G Guan H Li Y Luo F Zhao Q Zhang B Wei Q PeiH Peng Nat Photonics 9 233 (2015)
27 AR Hosseini CY Koh JD Slinker S Flores-Torres HD Abruntildea GG Malliaras ChemMater 17 6114 (2005)
28 H-C Su C-C Wu F-C Fang K-T Wong Appl Phys Lett 89 261118 (2006)29 H-C Su H-F Chen F-C Fang C-C Liu C-C Wu K-T Wong Y-H Liu S-M Peng
J Am Chem Soc 130 3413 (2008)30 JC de Mello N Tessler SC Graham RH Friend Phys Rev B 57 12951 (1998)31 JC de Mello JJM Halls SC Graham N Tessler RH Friend Phys Rev Lett 85 421
(2000)32 Q Pei G Yu C Zhang Y Yang AJ Heeger Science 269 1086 (1995)33 Q Pei Y Yang G Yu C Zhang AJ Heeger J Am Chem Soc 118 3922 (1996)34 DL Smith J Appl Phys 81 2869 (1997)35 JD Slinker JA DeFranco MJ Jaquith WR Silveira Y Zhong JM Moran-Mirabal H
G Craighead HD Abruntildea JA Marohn GG Malliaras Nat Mater 6 894 (2007)36 SV Reenen P Matyba A Dzwilewski RA J Janssen L Edman M Kemerink J Am
Chem Soc 132 13776 (2010)37 M Lenes G Garcia-Belmonte D Tordera A Pertegaacutes J Bisquert HJ Bolink Adv Funct
Mater 21 1581 (2011)38 K Lin LD Bastatas KJ Suhr MD Moore BJ Holliday JD Slinker ACS Appl
Mater Interfaces 8 16776 (2016)39 RD Costa E Ortiacute HJ Bolink F Monti G Accorsi N Armaroli Angew Chem Int Ed
51 8178 (2012)40 CH Lyons ED Abbas JK Lee MF Rubner J Am Chem Soc 120 12100 (1998)41 J-C Lepretre A Deronzier O Stephan Synth Met 131 175 (2002)
118 LD Bastatas and JD Slinker
42 M Armand F Endres DR MacFarlane H Ohno B Scrosati Nat Mater 8 621 (2009)43 ST Parker JD Slinker MS Lowry MP Cox S Bernhard GG Malliaras Chem Mater
17 3187 (2005)44 JD Slinker CY Koh GG Malliaras MS Lowry S Bernhard Appl Phys Lett 86
173506 (2005)45 MS Lowry JI Goldsmith JD Slinker R Rohl RA Pascal Jr GG Malliaras S
Bernhard Chem Mater 17 5712 (2005)46 RD Costa A Pertegas E Ortiacute HJ Bolink Chem Mater 22 1288 (2010)47 S Park D Moon S Damodharan M Chandran Y Choe Mater Res Bull 47 2807 (2012)48 CD Sunesh O Sunseong M Chandran D Moon Y Choe Mater Chem Phys 136 173
(2012)49 R Sun C-T Liao H-C Su Org Electron 15 2885 (2014)50 JD Slinker J Rivnay JA DeFranco DA Bernards AA Gorodetsky ST Parker M
P Cox R Rohl GG Malliaras S Flores-Torres HD Abruntildea J Appl Phys 99 074502(2006)
51 D Tordera S Meier M Lenes RD Costa E Ortiacute W Sarfert HJ Bolink Adv FunctMater 24 897 (2012)
52 Y Shen DD Kuddes CA Naquin T Hesterberg C Kusmierz BJ Holliday JD SlinkerAppl Phys Lett 102 203305 (2013)
53 KJ Suhr LD Bastatas Y Shen LA Mitchell BJ Holliday JD Slinker ACS ApplMater Interfaces 8 8888 (2016)
54 LD Bastatas K Lin MD Moore KJ Suhr MH Bowler Y Shen BJ Holliday JDSlinker Langmuir 32 9468 (2016)
55 SV Reenen RA Janssen M Kemerink Adv Funct Mater 22 4547 (2012)56 DR Blasini J Rivnay D-M Smilgies JD Slinker S Flores-Torres HD Abruntildea GG
Malliaras J Mater Chem 17 1458 (2007)57 KJ Suhr LD Bastatas Y Shen LA Mitchell GA Frazier DW Taylor JD Slinker B
J Holliday Dalton Trans 45 17807 (2016)58 SV Reenen P Matyba A Dzwilewski RAJ Janssen L Edman M Kemerink Adv
Funct Mater 21 1795 (2011)59 S Niu Z Cao T Yan J Phys Chem B 114 877 (2010)60 E Bandiello M Sessolo HJ Bolink J Mater Chem C 4 10781 (2016)
4 The Use of Additives in Ionic Transition Metal Complex hellip 119
Chapter 5Improving Charge Carrier Balanceby Incorporating Additives in the ActiveLayer
Hai-Ching Su
Abstract Light-emitting electrochemical cells (LECs) have recently drawn muchresearch interest because of their advantages related to single-layered devicearchitecture low-bias operation and employing inert electrodes Such featuresenable LECs to be inexpensive and efficient organic light-emitting devices Asmentioned in Chap 1 the lack of charge injection and transporting layers hinders toachieve balanced carrier mobilities in single-layered LECs In this chapter tech-niques reported in literatures to achieve charge carrier balance in LECs by incor-porating additives in the active layer are carefully reviewed Furthermore a newoptical technique to probe charge carrier balance in LECs is introduced to confirmthe effect attributed to the use of additives This chapter summarizes that improvingcharge carrier balance of LECs would be essential to realize high device efficiencyin spite of the existence of electrochemically doped layers
Keywords Charge carrier balance Carrier trap Carrier injection Carriertransport
51 Introduction
511 Characteristics of Light-Emitting ElectrochemicalCells (LECs)
Organic light-emitting diodes (OLEDs) have attracted intense scientific interestrecently owing to their useful applications for displays and solid-state lightingpurposes Nevertheless OLEDs suffered from some drawbacks such astime-consuming evaporation processes for multilayered device structures andemploying reactive metals as low-work-function electrodes To achieve simple
H-C Su (amp)Institute of Lighting and Energy Photonics National Chiao Tung UniversityTainan 71150 Taiwane-mail haichingsumailnctuedutw
copy Springer International Publishing AG 2017RD Costa (ed) Light-Emitting Electrochemical CellsDOI 101007978-3-319-58613-7_5
121
fabrication processes and to employ air-stable cathode materials light-emittingelectrochemical cells (LECs) were firstly demonstrated by Pei et al in 1995 [1] Ascompared to OLEDs LECs show some promising properties includingsingle-layered device structure low-bias voltage high power efficiency andcompatibility with air-stable cathodes In general LECs consist of only a singleorganic layer which is compatible with solution fabrication methods eg spincoating and inkjet printing Ionic species in LECs can induce electrochemicaldoping under a bias leading to low operating voltage even when a high carrierinjection barrier is present Balanced carrier injection under a low bias consequentlyresults in high power efficiency which is critical for lighting applicationsFurthermore since charge injection is insensitive to the work function of theelectrodes LECs can use inert electrodes eg gold and silver and thus can avoidcomplicated packaging processes
The active layer of an LEC has mobile ions which can move to electrodes undera bias Active materials of LECs are classified into four categories namelyfluorescent conjugated light-emitting polymers (CP) phosphorescent ionic transi-tion metal complexes (iTMC) small molecules and quantum dotsmdashsee Sect IIIfor more details For non-ionic emitters the active layers contain ionic salts to offermobile ions Ion-conducting polymers eg poly(ethylene oxide) (PEO) are oftenadded into the active layer to reduce phase separation between nonpolar polymersand polar salts For ionic emitters ion-conducting material is generally not requiredhowever ionic liquids eg 1-butyl-3-methylimidazolium hexafluorophosphate([BMIM][PF6]) are typically used as additives In addition electroluminescence(EL) efficiencies of iTMC-based LECs are commonly higher because of thephosphorescent nature of iTMCs
As explained in Sect 123 in Chap 1 two models were adopted to show theworking principle of LECs The first one is electrodynamical (ED) model andthe second one is electrochemical doping (ECD) model Mobile ions added in theactive layer of LECs move to electrodes under a bias and thus reduce carrierinjection barrier in both models However the reasons for reduction in carrierinjection barrier are different in these two models
In the ED model (left part of Fig 51) under a bias ions approach electrodesand electric double layers (EDLs) are formed leading to a significant drop of theelectric potential at the electrode interface Injected carriers move and recombine togenerate light in the field-free layer between the EDLs In the ECD model (rightpart of Fig 51) oxidized and reduced species form near anode and cathoderespectively Ohmic contacts at the electrode interfaces are established to enhancecarrier injection Doped layers penetrated into the intrinsic region at the center ofactive layer with time and a p-i-n structure is built up finally Electric potentialdecreases significantly in the intrinsic region and EL takes place because of carrierrecombination
Both models were evidenced by experiment results and simulation data Toclarify the working principle of LECs planar PLECs were employed since it is noteasy to measure the potential profile in the thin emissive layer (lt1 lm) of sandwichPLECs One of such planar PLECs was proposed by Pei et al [1] It was shown that
122 H-C Su
the light emission zone in the p-i-n junction located within a narrow areaFurthermore p- and n-doped layers in planar PLECs was also shown by capturingthe photoluminescence image and by measuring the electrostatic potential in planarPLECs via scanning Kelvin probe microscopy (SKPM) [2ndash5] Reported resultsshowed a deep potential drop in the center of the emissive layer [6 7] It agreedwell with the ECD model and many CP-based LECs showed similar behavior[8ndash10] Nevertheless the poly(phenylene vinylene) (PPV)-based planar LEC devicereported by Pingree et al showed that most potential (gt90) dropped at theemissive layercathode interface and only a small electric field was present acrossthe bulk of the emissive layer [11] Similar phenomena were found in planariTMC-based LECs [12] These data did not follow the ECD model well but agreedwith the ED model [13 14]
In 2010 van Reenen et al showed that using two different contacts ienon-injection limited and injection limited contacts in planar LECs lead to differentpotential profiles In non-injection limited ohmic contacts the devices agreed wellwith the ECD model forming a dynamic p-n junction in the bulk of the active layerOn the other hand in the injection-limited regime the device followed the EDmodel [15] Current-luminance-voltage (L-I-V) measurements were performed toreveal the working mechanism for sandwich LECs based on iTMCs [16] At a biasvoltage above the bandgap of the iTMCs ion separation led to formation of thep- and n-type doped layers and the central region was still neutral which corre-sponded to the ECD model Nevertheless at a bias below the bandgap of theiTMCs current transients exhibited the property following the ED model [16]These results indicated that the operation model of LECs depends on the injectionregimes If ohmic contacts are formed the LEC follows the ECD model while EDmodel fits well with experimental results when no ohmic contact is formed In case
Fig 51 Schematic diagrams of LECs based on electrodynamical (ED) model (left) andelectrochemical doping (ECD) model (right)
5 Improving Charge Carrier Balance by Incorporating Additives hellip 123
only one ohmic contact is formed only one type of doping occurs in the active layerand recombination zone is close to one of the electrodes
512 Charge Carrier Balance in LECs
Charge carrier balance is easily optimized in multilayered OLEDs using propercarrier injection andor transporting layers to adjust the imbalanced charge carriermobilities of the active layer Although balanced carrier injection in LECs is rel-atively easy to be achieved hole and electron mobility discrepancy of the emissivematerial would lead to an off-centered recombination zone Therefore excitonquenching near the doped layers reduces device efficiency [16] In general LECsare less sensitive to electrode work function since ohmic contacts for carrierinjection can be induced by the doped layers Nevertheless some literaturesrevealed that adjusting carrier injection efficiency shows pronounced effect oncarrier balance These results revealed that carrier balance of LECs would be alteredby modifying carrier injection ratios
For instance a self-assembled monolayer (SAM) accompanied with n-typedoping near the cathodepolymer interface was shown to offer a doubled increase indevice efficiency of CP-based LECs [17] The n-type doped SAM layer near theindium tin oxide (ITO) layer enhanced electron injection resulting in improveddevice efficiency These data showed that ohmic contact may not always true inLECs In 2004 Gorodetsky et al revealed how the electrode metals affect thedevice characteristics of LECs based on [Ru(bpy)3][PF6]2 (where bpy is22prime-bipyridyl) [18] Measured EL characteristics were insensitive to the electrodemetals employed under a forward bias (ITO wired as anode) Nevertheless differentanode metals injected holes at different efficiencies under a reverse bias Thethermal deposition damage at the organicmetal interface resulted in the differencein device performance between forward and reverse biases These results furtherreveal that adjusting carrier injection efficiency can modify carrier balance of LECs
Device performance of LECs can be improved by systematically adjusting thecharge carrier injection efficiency Liao et al demonstrated optimizing deviceefficiencies of LECs by employing proper carrier injection layers in 2012 [19] In astandard LEC device architecture [ITOpoly(34-ethylenedioxythiophene)poly(styrene sulfonate) (PEDOTPSS) (40 nm)emissive layer (200 nm)Ag (100 nm)]the peak external quantum efficiency (EQE) measured at 35 V was 852 Addinga thin hole injection layer (HIL) the hole injection efficiency enhanced and thecurrent flow was increased consequently Nevertheless the EQE decreased to676 It showed that improved hole injection lead to excess holes in the activelayer and in turn the carrier balance was deteriorated If an electron injection layer(EIL) was inserted in the standard LEC device both the device current and deviceefficiency increased It showed that holes are present as excess carriers in thestandard device and therefore increased electrons resulted in better carrier balanceThese results indicated that proper adjusting carrier injection efficiency would be
124 H-C Su
still essential for LECs to improve device efficiency in spite of electrochemicallydoped layers However employing carrier injection layers complicates fabricationprocesses and adjusting charge carrier balance in the emissive layer would be afeasible approach
513 Organization of this Chapter
In the following sections works focused on improving charge carrier balance byincorporating additives in the active layer are briefly reviewed Furthermore a newoptical technique to probe charge carrier balance in LECs is first introduced Itprovides direct evidence of shifted recombination zone affected by altered chargecarrier balance Then the techniques reported to optimize charge carrier balanceie to reach a centered recombination zone by incorporating additives in the activelayer are reviewed Finally conclusion and outlook are discussed
52 Optical Technique to Probe Charge CarrierBalance in LECs
To study the effect of incorporating additives on the charge carrier balance of LECsthe technology to probe the recombination zone position in the active layer is animportant work Nevertheless it is not easy to directly probe the recombinationzone in sandwiched LECs because of their thin active layer (generally lt1 lm) Oneof the alternative approaches is optically probing the recombination zone inlarge-interelectrode-gap (up to mm) planar LECs [2] However these data weremeasured in planar LECs with interdigitated electrode spacing much larger than theactive layer thickness of sandwiched devices Here the electric fields vary signif-icantly in magnitudes for planar and sandwiched LECs The ion mobility chargecarrier injection efficiency and carrier mobility which alter the recombination zoneposition are highly dependent on electric field Therefore the EL properties ofplanar LECs are not necessarily consistent with those of sandwiched LECs
Wang et al reported a new optical method to dynamically probe the recombi-nation zone position of sandwiched LECs via microcavity effect in 2013 [20]Although this technique is in depth described in Chaps 1 and 3 we provide a shortintroduction to assist the reader In short microcavity structures of sandwichedLECs alter the spectral outcoupling efficiencies and hence resulted in temporalevolution in EL spectra when the recombination zone moves Hence the temporalevolution in recombination zone position of sandwiched LECs can be obtained byfitting the time-dependent EL spectra to calculated EL spectra from microcavityeffect using proper emitting zone positions Outcoupled EL spectra of a bottomemitting LEC can approximately be simulated using the Eq 1 [21]
5 Improving Charge Carrier Balance by Incorporating Additives hellip 125
EextethkTHORNj j2frac14T2 1
N
PN
ifrac1411thornR1 thorn 2
ffiffiffiffiffiR1
pcos 4pzi
k thornu1
1thornR1R2 2ffiffiffiffiffiffiffiffiffiffiR1R2
pcos 4pL
k thornu1 thornu2
EintethkTHORNj j2 eth1THORN
where R1 and R2 are the reflectances from cathode and glass substrate respectivelyu1 and u2 are the phase changes of reflection from cathode and glass substraterespectively T2 is the transmittance from glass substrate L is the total opticalthickness of the active layers EintethkTHORNj j2 is the emission spectrum of the emissivematerials without microcavity effect EextethkTHORNj j2 is the outcoupled emission spectrumfrom glass substrate zi is the optical thickness between the emitting sublayer i andcathode The emitting zone is partitioned into N sublayers and the contributionsfrom all sublayers are summed up When p- and n-type doped layers were fullybuilt up the p-n junction width measured from capacitance measurements has beenreported to be ca 10 of the active layer thickness [22] In this manner theemitting zone width is estimated to be tenth of the emissive layer thickness Thephotoluminescence (PL) spectrum of the emissive layer on a quartz substrate isemployed as the emission spectrum without microcavity effect ( EintethkTHORNj j2) becauseno high-reflectance metal was in this sample By modifying the recombination zoneposition (zi) in the above shown equation the calculated EL spectra were fitted tothe measured one Fitted calculated and measured EL spectra extracted recombi-nation zone position of LECs
The proposed optical technique was performed on white LECs [23] Fittedcalculated and measured EL spectra along with consequently obtained recombi-nation zone position for white LECs exhibiting active layer thicknesses of 190 270and 400 nm are shown in Fig 52 For thinner devices (190 nm) shown in the leftpart of Fig 52 the steady-state recombination zone was closer to the anode and thedevice efficiency decreased due to exciton quenching from the p-type doped layerTherefore an EQE (75) which was similar to that of reported white LECscontaining the same complexes [24] was achieved Thickening the emissive layerthickness ensured sufficient spacing between the doped layers and the recombina-tion zone leading to reduced exciton quenching and enhanced device efficiency Asdepicted in the central part of Fig 52 the steady-state recombination zone inthicker devices (270 nm) was close to the central active layer and thus the peakEQE (power efficiency) was improved to ca 11 (20 lmWminus1) Further thickeningthe emissive layer thickness did not result in better device efficiency Thesteady-state recombination zone of thickest devices (400 nm the right part ofFig 52) was very near the anode and significant exciton quenching in the prox-imity of the p-type doped layer occurred again Thus much lowered EQEs (lt5)were obtained for this device Asymmetric recombination zone position in thethickest device may result from poorer charge carrier balance under a lower electricfield It reveals that the proposed optical technique is powerful in studying chargecarrier balance of LECs since definite recombination zone position can be extracted
126 H-C Su
53 Incorporating Carrier Trappers in LECs
Adding an efficient guest in a host layer is generally adopted in OLEDs to avoidself-quenching of guest molecules [25] Carrier trapping because of the offsets inenergy levels between host and guest molecules may simultaneously modify thecarrier balance Such phenomenon can be employed to improve charge carrierimbalance in the active layer Balanced amount of electrons and holes facilitates tomove the recombination zone toward the central active layer and hence it can avoidexciton quenching increasing device efficiency Su et al reported a double-dopingtechnique to improve device performance of white LECs in 2011 [24] Molecularstructures and energy levels of the complexes employed in this work are shown inFigs 53 and 54 respectively The blue-green-emitting complex (1 Fig 53) wasproposed by Tamayo et al [26] and the red-emitting complex (2 Fig 53) has beenutilized in the first white LECs based on iTMCs [27] Complex 1 LECs showed highEQEs up to 145 [28] revealing good carrier balance of complex 1 films whenphotoluminescence quantum yield (PLQY) of 75 [24] and ca 20 optical out-coupling efficiency are estimated Nevertheless the single-doped devices (1[BMIM][PF6]2 = 7982002 weight ratio) exhibited poor EQEs of 32 despite a goodPLQY of the active layer (61) [24] A larger energy level offset in the lowestunoccupied molecular orbital (LUMO) levels (046 eV) than that in the highestoccupied molecular orbital (HOMO) levels (039 eV) between host and guest is
400 500 600 700 80000
02
04
06
08
10 L = 190 nmzi = 110 nm
Measured SimulatedEL
Inte
nsity
(au
)
Wavelength (nm)
190 nm
110 nm
80 nm
400 500 600 700 80000
02
04
06
08
10 L = 270 nmzi = 140 nm
Measured SimulatedEL
Inte
nsity
(au
)
Wavelength (nm)
270 nm
140 nm130 nm
400 500 600 700 80000
02
04
06
08
10 L = 400 nmzi = 280 nm
Measured SimulatedEL
Inte
nsity
(au
)
Wavelength (nm)
400 nm
280 nm120 nm
Fig 52 Simulated and measured stabilized EL spectra of white LECs with active layerthicknesses of 190 (left) 270 (central) and 400 nm (right) Recombination zone position (zi)obtained from fitting of calculated and measured EL spectra is labeled in the inset of eachsubfigure
1
Ir
N
NN
N
+F
F
F
F
N
N
[PF6]-
2
Ir
N
N
+
N
N
[PF6]-
3
Ir
N
N
+
N
N
[PF6]-
Fig 53 Molecular structures of the 1 2 and 3 emitting iridium(III) complexes used in Ref [24]
5 Improving Charge Carrier Balance by Incorporating Additives hellip 127
present in single-doped active layer (Fig 54) This would result in more significanttrapping of electron and hence carrier balance of the emissive layer deteriorates If anorange-emitting guest (3 Fig 53) showing a higher LUMO level (Fig 54) wasadditionally doped in the emissive layer to mitigate electron trapping carrier balancecan be better The double-doped white LECs (1[BMIM][PF6]23 = 79852000501 weight ratio) exhibited superior EQE (power efficiency) up to 74(15 lmWminus1) [24] Such data reveal that the double-doping method to modify carriertrapping effect is a feasible technique to improve carrier balance of LECs
Similar concept was employed by Liao et al who proposed improving carrierbalance of hostndashguest LECs containing an ionic terfluorene host (4 Fig 55) dopedwith a red-emitting iridium(III) complex (2 Fig 53) [29] As compared to iTMChosts cheaper fluorescent emissive materials are easy to independently tailor carriertransporting properties and energy gaps Hence it is more suitable to be a hostmaterial in LECs Compound 4 exhibited yellow phosphorescent photolumines-cence (PL) with a peak at 562 nm which corresponds to an excited triplet stateenergy of 221 eV [30] Therefore it is suitable to serve as the host material for ared-emitting guest Terfluorene derivatives containing alkyl substitutions on thetetrahedral C9 carbon were shown to have higher electron mobilities [31]Furthermore imidazole moieties which contain alkyl chains on compound 4 wereutilized in electron-transport materials for OLEDs [32] Thus electron preferredtransport properties of 4 possibly result in poor device efficiency of the host-onlyLECs (EQE 1) [29] If a low-gap guest which leads to electron trapping isadded in 4 host electron transporting is impeded and charge carrier balance isimproved The energy levels of the host 4 and the guest 2 are shown in Fig 56
2
-566 eV
-343 eV
1
-605 eV
-297 eV -304 eV
-568 eV
3
Fig 54 Energy levels of theblue (1) red (2) and orange(3) emitting iridium(III)complexes used in Ref [24]
4
N N
N N
+2
[PF6]2-
Fig 55 Chemical structureof the ionic terfluorene host(4) employed in Ref [29]
128 H-C Su
Both molecules exhibit similar HOMO level energies but they show a large energyoffset in the LUMO levels These energy level alignments result in electron trappingcompensating the carrier mobility imbalance in the host material When doped with aguest of 05 wt the peak EQE of the hostndashguest LECs was up to 362 Thisapproached the upper limit (4) estimated from the PLQY of the guest doped atlow concentrations in the host (20) when estimating spin statistics 100(both singlet and triplet excitons are emissive for phosphorescent material) and anoptical outcoupling efficiency of 20 Such data demonstrated that good chargecarrier balance of the hostndashguest LECs is realized by adding a guest with properHOMOLUMO levels to adjust the charge carrier mobilities in the host film
Doping a low-gap guest to be a charge carrier trapper would improve the chargecarrier balance of LECs Liao et al showed improving charge carrier balance inLECs by using 1 (Fig 53) as the host material and the cationic near-infrared(NIR) laser dye 33prime-diethyl-22prime-oxathiacarbocyanine iodide (DOTCI) as a chargecarrier trapper [33] Undoped complex 1 LECs exhibited an EQE 9 which wassignificantly lower than the maximum (15) expected from the PLQY of theactive layer (75) and an optical outcoupling efficiency of 20 [24] Thisimplied poor carrier balance of the undoped LECs Nevertheless with DOTCI as ahole trapper the device efficiency increased significantly (EQE 1275) Thedoping concentration of the hole trapper was very low (001 wt) and thus theenergy transfer efficiency was low rendering almost identical host-only EL emis-sion as compared to that of the undoped LECs Enhanced device efficiency becauseof hole trapper (Fig 57) doping indicated that holes are more than electrons in the
4
-567 eV
-211 eV
-566 eV
-343 eV
2
Fig 56 The host (4) and theguest (2) energy levels inRef [29]
Fig 57 The host (1) and theguest (DOTCI) energy levelsin Ref [33]
5 Improving Charge Carrier Balance by Incorporating Additives hellip 129
undoped LEC Another work also reported that increased electron injection lead toimproved device efficiency of LECs based on 1 confirming that hole is the excesscarrier [19] It is useful for LECs based on iTMCs because the guest concentrationfor adjusting charge carrier balance is rather low and device performance can beoptimized without changing the EL spectrum
The proposed optical technique mentioned in Sect 52 can be employed tostudy the effect of carrier trapper on charge carrier balance of the LEC [20]Commercially available [Ru(dtb-bpy)3][PF6]2 (where dtb-bpy is 44prime-ditertbutyl-22prime-bipyridine) (5) was used in this study [34] NIR laser dye33prime-diethylthiatricarbocyanine iodide (DTTCI) was used as the carrier trapper inthe active layer (450 nm) of 5 After a bias of 25 V was applied to the neat-filmLECs the recombination zone position (zi) from the anode at 8 12 18 and 58 minwere 100 150 230 and 250 nm respectively Under a bias the required number ofions at the anode to reach ohmic contact for hole was less than that required to reachohmic contact for electron at the cathode because the hole injection barrier waslower (Fig 58) Thus the hole injection efficiency was higher initially and therecombination zone was closer to the cathode (zi = 100 nm) After both p- andn-type doped layers were well built up carrier injection was getting balanced andthe recombination zone approached the central emissive layer (zi = 250 nm)Nevertheless temporal moving direction of the recombination zone was oppositewhen doped with the hole trapper DTTCI The recombination zone was initiallynear the anode (zi = 390 nm) and then moved toward the central active layer(zi = 295 nm) This indicated that p- and n-type doped layers have not completelyestablished yet shortly after a bias was just applied and hence the recombinationzone position was significantly affected by hole trapping from low-gap guest(Fig 58) When the doping processes proceeded enhancement rate in holeinjection was faster than that in electron injection because of relatively smaller holeinjection barrier Therefore the recombination zone shifted toward the centralactive layer gradually These results provide clear evidence to prove that doping acarrier trapper affected charge carrier balance and hence it is a useful method toimprove LEC performance
Fig 58 Energy levels of 5and the hole trapper DTTCIalong with the electrode workfunctions used in Ref [20]
130 H-C Su
54 Incorporating Salts in LECs
Salts were often used in LECs to offer more mobile ions speeding the deviceresponse upmdashsee Chap 4 for more details Besides improved ionic conductivitycharge carrier injection and transporting characteristics of the active layer weresimultaneously altered Positions of recombination zone in CP-based LECs werereported to be dependent on the cation size of the polymer electrolytes [35] Theplanar LECs containing larger cations lead to emission zones closer to the centralactive layer (rubidium or cesium) while devices with smaller cations resulted inoff-centered emission zones (lithium sodium and potassium) [35] Centeredemission zones caused better device efficiencies due to mitigated exciton quenchingin the proximity to the doped layers Another research work also indicated similarresults [36] These results were resulted from ionic mobility and electrochemicaldoping speed in the emissive layer Although adjusting carrier injection andtransporting characteristics were not attributed to the cations these works impliedthe possibility of adjusting charge carrier balance by adding salts with various sizes
Salts were used to improve brightness efficiency and response time of LECsbased on [Ir(ppy)2(bpy)][PF6] (where ppy is 2-phenylpyridine)mdashsee Chap 7 formore details about this compound [37] When doped with the lithium salt the LECsexhibited gt4X improvement in maximum luminance and EQE In addition lithiumsalts resulted in a tenth of device turn-on time Enhanced luminance and efficiencyof LECs revealed that better charge carrier balance was obtained by adding salt Thecation radius of the [Ir(ppy)2(bpy)]
+ (63 Aring) is larger than that of the [PF6]minus
counterion (16 Aring) Thus the cation packing density at the cathode was lower thananion cation packing density at the anode [37] Electron injection efficiency wasconsequently relatively lower in iTMC LECs Increasing smaller cations lead toenhanced cation packing density at the cathode improving the charge carrier bal-ance It clearly demonstrated that adding proper salts improves the device effi-ciencies of LECs
In 2014 Sun et al showed modifying charge carrier balance of iTMC LECs bydoping salts containing different alkyl chain lengths [38] Imidazole-based salts
N N R
[EMIM]+ R=C2H5[BMIM]+ R=C4H9[HMIM]+ R=C6H13[OMIM]+R=C8H17
[PF6]-+Fig 59 Chemical structures
of the imidazole-based saltsemployed in Ref [38]
5 Improving Charge Carrier Balance by Incorporating Additives hellip 131
having various alkyl chain lengths were doped in the active layer of LECs based on5 The salts used in this work are 1-ethyl-3-methylimidazolium hexafluorophos-phate ([EMIM][PF6]) 1-butyl-3-methylimidazolium hexafluorophosphate ([BMIM][PF6]) 1-hexyl-3-methylimidazolium hexafluorophosphate ([HMIM][PF6]) and1-octyl-3-methylimidazolium hexafluorophosphate ([OMIM][PF6]) (Fig 59) Thecurrent density reduced with the length of alkyl chain while doubled device effi-ciency was obtained when a proper salt was used To further study the effect of thealkyl chain length of salt on the device characteristics temporal evolution of therecombination zone upon adding different salts was estimated by the techniqueshown in Sect 52 Estimated temporal recombination zone for devices with nosalt [EMIM][PF6] [BMIM][PF6] [HMIM][PF6] and [OMIM][PF6] are shown inFig 510 Devices containing [BMIM][PF6] and [HMIM][PF6] exhibited relativelycentered recombination zones showing EQEs higher than 3 The recombinationzone of [EMIM][PF6] device was closer to the cathode while that of [OMIM][PF6]device was closer to the anode They suffered significant exciton quenching in theproximity to the doped layers and in turn they exhibited lower device efficiencyNon-doped device showed the most asymmetrical recombination zone which wasvery near the anode resulting in the lowest EQE lt2 It indicated that a long alkyl
600 nm
492 nmNo salt
402 nmAnod
e
Cath
ode
264 nmEMIMPF 6168 nmAn
ode
Cath
ode
324 nmBMIMPF6492 nmAn
ode
Cath
ode
360 nmHMIMPF6480 nmAn
ode
Cath
ode
444 nmOMIMPF6558 nmAn
ode
Cath
ode
Fig 510 Temporalrecombination zone evolutionof the LECs containing 5added with no salt [EMIM][PF6] [BMIM][PF6][HMIM][PF6] and[OMIM][PF6]
132 H-C Su
chain length of salt may hinder hole transporting more significantly than electrontransporting Therefore device performance of LECs can be improved by adding asalt with a proper alkyl chain length to reach centered recombination zone whichreduced exciton quenching
55 Incorporating Carrier Transport Materials in LECs
Charge carrier balance of LECs would be improved by mixing carrier transportingmaterials in the emissive layer Chen et al reported LECs containing a triarylaminederivative as the hole-transport material (HTM) and a 135-traizine derivative asthe electron-transport material (ETM) [39] The chemical structures of HTM (6) andETM (7) employed in this work are shown in Fig 511 Both compounds 6 and 7were deep-blue-emitting materials while a green emission with a peak at 545 nm inthe mixing film was observed Such green emission was reduced by doping the inertpolymer poly(methylmethacrylate) (PMMA) showing that the green emission wasattributed to the 67 exciplex The EL spectrum of the LECs based on mixed 67films was almost independent on the mixing ratios and was similar to the PLspectrum of the emissive layer Nevertheless the device efficiency was highlyrelated to the mixing ratios With 25 wt ETM 7 doped in HTM 6 LECs 10Ximprovement in EQE was reached Another ten-fold enhancement in EQE wasobtained by increasing the ETM 7 concentration to 50 wt Eventually the peakEQE of the LEC containing 75 wt ETM 7 reached ca 26 This represents anincrease factor of 350 in the device efficiency It confirmed that highly efficientLECs would be achieved by adjusting charge carrier balance of the active layer
According to a similar idea Pertegaacutes et al showed hostndashguest LECs containingan ionic HTM and a neutral ETM [40] The guest material was a neutralblue-emitting complex bis(46-difluorophenylpyridinato-NC2prime)picolinate (FIrpic)
6 7
Fig 511 Chemical structures of the hole-transporting material (6) and electron-transportingmaterial (7) employed in Ref [39]
5 Improving Charge Carrier Balance by Incorporating Additives hellip 133
The performance of the hostndashguest LECs was dependent on the HTMETM mixingratio The device efficiency was shown to be maximized by adjusting theHTMETM weighting ratio revealing improved charge carrier balance
Recently charge carrier balance of LECs containing 5 has been improved bydoping a polymer HTM into the active layer [41] The molecular structure of HTM(8) used in this study is shown in Fig 512 Carbazolyl moieties enhanced holetransport and alkoxy groups avoided phase separation from the ionic complex hostWhen the p- and n-type doped layers in LECs were well established narrowedintrinsic layer resulted from extending doped layers rendering exciton quenching inthe asymmetrical recombination zone As such the device efficiency reduced sig-nificantly Doping a polymer HTM in the active layer lead to a central recombi-nation zone reducing exciton quenching Estimating time-dependent recombinationzone in LECs containing a polymer HTM was performed by using the methodmentioned above As shown in the upper part of Fig 513 the initial (30 min)recombination zone in non-doped device was rather approaching anode With 7HTM 8 doped in the active layer the recombination zone shifted toward the centralactive layer (the central part of Fig 513) and the device EQE increased signifi-cantly (from 137 to 229) Recombination zone moving was attributed to dopingHTM which enhanced hole transporting Central recombination zone lead toenough spacing between the recombination zone and the doped layers renderingreduced exciton quenching and enhanced device efficiency consequentlyIncreasing the HTM 8 concentration up to 12 shifted the recombination zonecloser to the central active layer (the lower part of Fig 513) It resulted in a furtherincreased EQE (248) The EQE was increased by 80 by doping 12 HTM 8 Itrevealed that doping a polymer HTM into the active layer of LECs pushed therecombination zone away from anode Thus exciton quenching was reduced andhigher light output and device efficiency was reached
8
Fig 512 Molecularstructure of the hole-transportmaterial (8) used in Ref [41]
134 H-C Su
56 Conclusion and Outlook
Since the first work of LECs in 1995 [1] many research works on emissivematerials for LECs have been reported to improve material PLEL propertiesrendering device efficiency approaching that measured in OLEDs Neverthelesshigh PLQY of emissive materials did not necessarily ensure high EL efficiency inLECs since superior charge carrier balance was also required This was a commonproblem for most of the reported LECs In this chapter optical technique to probecharge carrier balance in LECs has briefly been introduced In addition severalreported methods to provide better charge carrier balance of LECs were reviewedmdashsee Chaps 1 and 3 for more details about this technique Carrier trapping due toenergy level offset between host and guest can modify the charge carrier mobilitiesimproving the charge carrier balance of LECs Incorporation of salts into theemissive layer also affected the charge carrier mobility of the emissive materialTherefore to choose proper salts could also lead to better charge carrier balance ofLECs Incorporating carrier transport materials in LECs would be another feasibleway to adjust charge carrier mobility of the active layer Enormous improvement indevice efficiency was also realized by means of balancing the charge carriermobilities Overall these works revealed that even with the p- and n-type dopedlayers adjusting charge carrier balance of LECs was required to achieve betterdevice performance In the future developments in functional additives may beessential to further improve charge carrier balance of LECs Such functional
Emissive layer
PED
OT
PSS
ITO
2 nm
60 nm
570 nmAg
125 nm
155 nm
HTM 0
HTM 7
HTM 12
60 nm
125 nmPE
DO
TPS
S
ITO
PED
OT
PSS
ITO
Emissive layer Ag
Emissive layer Ag
Fig 513 Temporal recombination zone evolution of the LECs containing 5 doped with HTM 8of 0 (upper) 7 (central) and 12 (lower)
5 Improving Charge Carrier Balance by Incorporating Additives hellip 135
additives have to be composed of charge-transport fragments with ionic groups toprevent phase separation Finally charge-transport characteristics of functionaladditives can be adjusted by modifying the molecular structures Functional addi-tives can be further included in the emissive materials of LECs to simplify thefabrication processes Improving charge carrier balance of LECs will be a chal-lenging work and continuous efforts should be made to achieve better deviceefficiencies
Acknowledgements The author acknowledges the financial support from Ministry of Scienceand Technology (MOST 105-2221-E-009-097-MY2)
References
1 Q Pei G Yu C Zhang Y Yang AJ Heeger Science 269 1086 (1995)2 J Gao J Dane Appl Phys Lett 83 3027 (2003)3 J Dane C Tracy J Gao Appl Phys Lett 86 153509 (2005)4 J Dane J Gao Appl Phys Lett 85 3905 (2004)5 JH Shin P Matyba ND Robinson L Edman Electrochim Acta 52 6456 (2007)6 P Matyba K Maturova M Kemerink ND Robinson L Edman Nat Mater 8 672 (2009)7 DB Rodovsky OG Reid LS Pingree DS Ginger ACS Nano 4 2673 (2010)8 L Edman Electrochim Acta 50 3878 (2005)9 Q Sun Y Li Q Pei IEEEOSA J Disp Technol 3 211 (2007)
10 C Yang Q Sun J Qiao Y Li J Phys Chem B 107 12981 (2003)11 LSC Pingree DB Rodovsky DC Coffey GP Bartholomew DS Ginger J Am Chem
Soc 129 15903 (2007)12 JD Slinker JA Defranco MJ Jaquith WR Silveira YW Zhong JM Moran-Mirabal
HG Craighead HD Abruntildea JA Marohn GG Malliaras Nat Mater 6 894 (2007)13 Q Pei AJ Heeger Nat Mater 7 167 (2008)14 Y Hu J Gao J Am Chem Soc 133 2227 (2011)15 SV Reenen P Matyba A Dzwilewski RAJ Janssen L Edman M Kemerink J Am
Chem Soc 132 13776 (2010)16 M Lenes G Garcia-Belmonte D Tordera A Pertegaacutes J Bisquert HJ Bolink Adv Funct
Mater 21 1581 (2011)17 DT Simon DB Stanislowski SA Carter Appl Phys Lett 90 103508 (2007)18 AA Gorodetsky S Parker JD Slinker DA Bernards MH Wong GG Malliaras
S Flores-Torres HD Abruntildea Appl Phys Lett 84 807 (2004)19 CT Liao HF Chen HC Su KT Wong Phys Chem Chem Phys 14 9774 (2012)20 TW Wang HC Su Org Electron 14 2269 (2013)21 X Liu D Poitras Y Tao C Py J Vac Sci Technol 22 764 (2004)22 IH Campbell DL Smith CJ Neef JP Ferraris Appl Phys Lett 72 2565 (1998)23 YP Jhang HF Chen HB Wu YS Yeh HC Su KT Wong Org Electron 14 2424
(2013)24 HC Su HF Chen YC Shen CT Liao KT Wong J Mater Chem 21 9653 (2011)25 CW Tang SA VanSlyke CH Chen Appl Phys Lett 65 3610 (1989)26 AB Tamayo S Garon T Sajoto PI Djurovich IM Tsyba R Bau ME Thompson
Inorg Chem 44 8723 (2005)27 HC Su HF Chen FC Fang CC Liu CC Wu KT Wong YH Liu SM Peng J Am
Chem Soc 130 3413 (2008)28 HB Wu HF Chen CT Liao HC Su KT Wong Org Electron 13 483 (2012)
136 H-C Su
29 CT Liao HF Chen HC Su KT Wong Phys Chem Chem Phys 14 1262 (2012)30 HF Chen CT Liao TC Chen HC Su KT Wong TF Guo J Mater Chem 21 4175
(2011)31 WY Hung TH Ke YT Lin CC Wu Appl Phys Lett 88 064102 (2006)32 AP Kulkarni CJ Tonzola A Babel SA Jenelhe Chem Mater 16 4556 (2004)33 CT Liao HF Chen HC Su KT Wong J Mater Chem 21 17855 (2011)34 S Bernhard JA Barron PL Houston JL Ruglovksy X Gao GG Malliaras J Am
Chem Soc 124 13624 (2002)35 Y Hu J Gao Appl Phys Lett 89 253514 (2006)36 JH Shin ND Robinson S Xiao L Edman Adv Funct Mater 17 1807 (2007)37 Y Shen DD Kuddes CA Naquin TW Hesterberg C Kusmierz BJ Hollday JD
Slinker Appl Phys Lett 102 203305 (2013)38 R Sun CT Liao HC Su Org Electron 15 2885 (2014)39 HF Chen CT Liao HC Su YS Yeh KT Wong J Mater Chem C 1 4647 (2013)40 A Pertegaacutes NM Shavaleev D Tordera E Orti MK Nazeeruddin HJ Bolink J Mater
Chem C 2 1605 (2014)41 PC Huang G Krucaite HC Su S Grigalevicius Phys Chem Chem Phys 17 17253
(2015)
5 Improving Charge Carrier Balance by Incorporating Additives hellip 137
Chapter 6Morphology Engineering and IndustrialRelevant Device Processingof Light-Emitting Electrochemical Cells
G Hernandez-Sosa AJ Morfa N Juumlrgensen S Tekogluand J Zimmermann
Abstract This chapter discusses the importance of the interplay between materialsmorphology processing and device performance for the fabrication oflight-emitting electrochemical cells (LEC) by industrially relevant technologiesThis is centered on the utilization of the polymer solid electrolytes as a mean to tunedevice performance film morphology and rheological properties Tunableparameters include the choice of molecular weight material ratio or monomer ratioin the case of copolymers We highlight the advantages of LECs over other tech-nologies in terms of their simplicity of fabrication by reviewing the latest researchon printed devices utilizing techniques like gravure and inkjet printing as well asspray and slot-die coating
Keywords Light-emitting electrochemical cells Solid electrolyte Printingparameters Layer morphology Industry-relevant printing methods
61 Introduction
Organic light-emitting diodes (OLED) represent the leading device architecture inwhich basic research in the organic electronic field has managed to translate lab-oratory prototypes into marketable products [1] High-end electronic products suchas mobile phones and televisions now contain displays based on OLED technology[2] The fabrication of highly efficient OLEDs for these applications usuallyrequires a multilayered device stack with controlled thicknesses generally involvingseveral costly vacuum evaporation steps [3 4] This process does not comply withthe promise of organic electronics in terms of low-cost and high-volume production
G Hernandez-Sosa (amp) AJ Morfa N Juumlrgensen S Tekoglu J ZimmermannLight Technology Institute Karlsruhe Institute of Technology Engesserstrasse 13Karlsruhe 76131 Germanye-mail gerardososakitedu
G Hernandez-Sosa AJ Morfa N Juumlrgensen S Tekoglu J ZimmermannInnovationLab Speyererstrasse 4 Heidelberg 69115 Germany
copy Springer International Publishing AG 2017RD Costa (ed) Light-Emitting Electrochemical CellsDOI 101007978-3-319-58613-7_6
139
of a wide range of applications such as packaging wearable or signage In contrastsolution processing of optoelectronic devices particularly through printing orcoating techniques is set to become the de facto fabrication choice for thecost-effective production of electronics [5] This is due to the advantages likehigh-throughput reduced material waste etc that this technique has delivered forgraphical printing However the complexity of OLEDs with the inherent solubilityand intermixing issues of its multilayer architecture along with the tight tolerancesnecessary for film thicknesses (lt10 nm) and defects will hinder a low-cost solutionfor the fabrication of fully printed devices in a continuous process
Light-emitting electrochemical cells (LEC) represent an alternative illuminatingdevice with very distinct processing advantages over OLEDs [6ndash9] as well as beinga competitive cost-effective technology [10] As commented in Chap 1 the sim-pler device architecture of a LEC compared to that of an OLED is presented inFig 61 It consists of a single active layer placed between two electrodes with norestriction to their work function The active layer can be composed of a blend of alight-emitting material (ie polymer small molecule and quantum dots seeChaps 10 12 and 13) and a polymer solid electrolyte alternatively all of the ionicand electronic functions can be combined in a single component through the use ofan ionic transition-metal complex (iTMC) blended with ionic liquids (see Chaps 47 8 and 11) [11 12] The working principle of LECs is based on the dynamicformation of a p-i-n light-emitting junction enabled by the existence of ionic specieswithin the single active layer With applied bias the ionic species in the active layermigrate to the corresponding electrode interface forming electric double layers(EDL) These EDLs enable charge carrier injection from the electrodes In accor-dance with the electrochemical doping model the injected carriers (ie electrons orholes) oxidize or reduce the semiconductor material [9] These redox processes leadto the movement of the corresponding ionic species to the redox centers to stabilizeand electrostatically compensate the locally oxidized or reduced semiconductorThis process results in an electrochemically doped semiconductor Finally a narrowintrinsic region is left unreacted between the doped zones where the formation ofexcitons and its subsequent radiative recombination takes place [13 14] Moredetails about the device mechanism are provided in Chap 1
Fig 61 Scheme of device architectures of an LEC (left) and an OLED (right)
140 G Hernandez-Sosa et al
In contrast to OLEDs a single active layer combines the function of the nec-essary injection and emitting layers thus reducing fabrication process steps Themain consequence of the dynamic formation of the charge injection layers is that nolow-work function metal (ie Ca LiF or Ba) is necessary for the injection ofelectrons from air stable electrodes This feature bypasses one of the main chal-lenges for fully solution processing OLEDs [15 16] and also allows for the fab-rication of metal free devices as presented in Fig 62 [6]
In addition to the advantageous working principle LECs with active layerthicknesses of up to 1 microm have been reported highlighting a large tolerance tothickness variations [17] This is a great advantage since printing of functionalfluids is not directly comparable to graphical printing This means that after filmdeposition the layer quality must be of sufficient quality to ensure the functionalityof the device in contrast to merely displaying information by color patternsFurthermore each different printing technique requires a specific window of rhe-ological properties of the printing fluid in order to yield an adequate film suitablefor device fabrication Simultaneously each printing technique presents differentcharacteristics in terms of printing speed and lateral and vertical resolution that willdefine its use according to the final application (ie pixels electrical contacts etc)[18] In this context the search for an optimum device architecture with a definedcomposition or morphology that yields the highest performance is as important asthe control of the physical properties that govern a certain deposition techniqueprocess Therefore the formulation of inks and study of the process parametersnecessary to achieve reproducible devices needs to be simultaneously studied toachieve the best compromise between performance and processability The recentadvances in the study of thin film processing and the fabrication of LECs by
Fig 62 All organic LEC Reprinted with permission from Ref [6] Copyright 2010 AmericanChemical Society
6 Morphology Engineering and Industrial Relevant Device hellip 141
industry-relevant techniques highlighted herein aim to enable the low-costhigh-throughput production and the eventual commercialization of items such asflexible displays smart packaging luminescent wall-papers etc where the uniquecharacteristics of these devices can be taken to their full potential
62 Film Morphology and Polymer Solid Electrolytes
621 Introduction
LECs have generally been reported to exhibit a lower performance than OLEDs interms of lifetime brightness and turn-on time This is caused by several factorssuch as electrochemical instability of the material system imbalanced ionicmobility or phase segregation of the components (see Chaps 5 7 8 and 12 formore details) [7 19] In the literature polymer-based LECs have commonly beenfabricated with poly (phenyl vinylene) (PPV) derivatives and a poly (ethyleneoxide) (PEO)-based electrolyte while iTMCs have been based on iridium(III)complexes sometimes mixed with an ionic liquid Recent reports have presenteddevices with operational lifetimes over several thousand hours at 10 cdA and over4000 h at 600 cdm2 for polymer- and iTMC-based LECs respectively [19 20]These encouraging results motivate the investigation of the specific detrimentalelectrochemical side reactions within LECs as well as the search for optimummorphologies in promising material systems
The properties of the electrolyte component of the active layer greatly determinethe properties of the film morphology device performance and stability Materialssuch as ionic liquids polymersndashsalt blends polyelectrolytes and cross-linkablecompounds have been used (see Chaps 4 and 5) [21] with all of the concomitantadvantages and disadvantages in terms of processability and physical and electro-chemical properties In this section we will focus on polymer-based solid elec-trolytes (PSE) Due to their polymeric nature changes in molecular structure willnot only affect film morphology but can also have an influence in the rheologicalproperties of the active layer blend [22ndash24] This property is of the upmostimportance to the film processing by printing techniques since it offers a broadtoolbox to obtain suitable ink formulations for the different printing techniqueswhile keeping control of the film morphology The latter is usually investigatedthrough spin coating For example changing the molecular weight of a polymer canlead to changed ionic transport properties solubility or rheology as will be shownherein [23] A second example highlighted in this chapter is the use of a copolymerbased PSE where the influence of the monomer ratio was investigated The char-acterization methods utilized in these studies are optical and atomic force micro-scopy (AFM) fluorescence microscopy as well as viscosity measurements
142 G Hernandez-Sosa et al
622 LEC Active Layer Morphology PSE Phase Separation
Matyba et al performed a study that exemplifies the relation between the activelayer components and its correlation to morphology electrochemical stability anddevice performance [24] They investigated a series of thin films and devicescomprising a PSE blend of PEO and KCF3SO3 with several PPVs based semi-conductors Optical micrographs of blends of PEO with the different conjugatedpolymers are presented in Fig 63 It was shown that the three materials phaseseparate into ldquodarkerrdquo (light absorbing) conjugated polymer areas and a ldquobrighterrdquoPSE phase The surface morphology was also reported to be dependent on the typeof conjugated polymer used MEH-PPV-based blends exhibit a comparatively smalldomain size while the BUEH-PPV and BEHP-co-MEH-PPV based blends showlarger domain sizes This is in line with previously performed studies on suchmaterials [25 26] The effect of domain size on device performance was alsoinvestigated Smaller domain sizes were shown to correspond to a higher ionicconductivity Ultimately device performance was limited by the electrochemicalwindow of the PSE system in relation to the redox potentials of the differentsemiconductors highlighting the key need of improving the electrolyte forimproved operation of LECs
623 Effect of the PSE Molecular Weight on the FilmMorphology
In our previous work we investigated the effect of the molecular weight (Mw) of thepolymeric component of the PSE on the morphology performance and rheologicalproperties of the active layer material [23] Here we used a PSE based on poly(methyl methacrylate) (PMMA) with Mw ranging from 3500 to 1000000gmol The PSE system was completed with the inclusion of [TBA][BF4] as the ionsource A standard ratio of 1125025 was used for the active layer componentsthe semiconductor Super Yellow (SY) PMMA and [TBA][BF4] We investigatedthe morphology by means of AFM and fluorescence microscopy The surfacemorphology of the fabricated devices is shown in the AFM micrographs presentedin Fig 64a For samples with a Mw higher than 3500 round-shaped domains(P) and networking paths (S) were observed hinting at phase separation betweenthe PSE and the semiconductor SY Elongation of the P domains was observed toscale with increasing Mw of PMMA Such phase separation occurred by a spinodaldecomposition mechanism originating from the entropy gain in polymerpolymerblends [27 28] The root mean square roughness within each of the P and Sdomains was found to be below 08 nm and 14 nm respectively
6 Morphology Engineering and Industrial Relevant Device hellip 143
Fig 63 Optical micrographs of conjugated polymerPSE blends showing phase separationReprinted from Ref [24] Copyright 2008 with permission from Elsevier
144 G Hernandez-Sosa et al
Fig 64 AFM (top) and optical microscopy images (bottom) of the LECs active layers with thedifferent PMMA Mw a 3500 gmol b 120000 gmol c 350000 gmol and d 996000 gmol TheLECs microscope images were taken in operation at a constant current density (15 mAcm2)Reprinted from Ref [23] Copyright 2013 with permission from Elsevier
6 Morphology Engineering and Industrial Relevant Device hellip 145
Figure 64b presents optical micrographs of LECs under operation with fourdifferent PMMA Mw in the PSE layer The devices were operated at a constantcurrent density of nominally 15 mAcm2 and were observed to exhibit an inho-mogeneous light emission with bright and dark areas throughout the sample Thedevice comprising a PSE with PMMA with a Mw of 3500 also showed an inho-mogeneous emission as a result of phase separation Surprisingly this phase sep-aration was not evident by AFM The correlation of the AFM measurements and theoptical micrographs demonstrate that only the S domains are electroluminescentTherefore they must contain the light-emitting SY as well as ionic species Theinset on each image corresponds to a fluorescence micrograph of the same area inthe sample It can be seen that only the S domains present light emission sug-gesting that the P domains are either SY poor or only comprised of the PSE As aconsequence of the existence of only one electroactive phase the overall effectiveactive area of the devices was reduced Besides phase separation theMw of the PSEis known to affect its ionic conductivity In turn it has also been found in literaturethat the turn-on voltage (Von) and turn-on time of a LEC correlates directly with theintrinsic conductivity of the PSE [19 29 30]
The current-voltage characteristics of the fabricated devices are presented inFig 65 It was observed that the devices prepared with a PMMA Mw of 350000have the lowest Von as well as the highest luminance (ie efficiency) The PSEionic conductivity was determined to be 94 10minus9 Scm 12 10minus8 Scm29 10minus7 Scm and 13 10minus7 Scm for the 3500 120000 350000 and1000000 Mw respectively The ionic conductivity of the PSE was found to bedirectly correlated to the LEC Von which we defined as the voltage at which aluminance value of 1 cdm2 is measured Although it was not experimentallydetermined the turn-on time should also be effected in the same way by the ionicconductivity [29] The observed dependence of the Von at a constant sweep rateshould correspond to a reduction of the turn-on time The inset in Fig 65b shows aphotograph of a device prepared with a PMMA Mw of 350000 under operation Inspite of the phase separation a homogeneous emission was perceived throughoutthe whole device area However when accounting only for the effective active areaof light emission as observed in Fig 64 the effective luminance per area increasesby a factor 40 for all cases
Interestingly the change of the PSEMw can also be used to simultaneously affectthe viscosity of the active layer solution This is because the Mw of a polymer isdirectly proportional to the fluid viscosity at a given concentration Figure 66shows that fluids composed of solvent SYPMMA[TBA][BF4] demonstrate strongshear-thinning at shear rates above 100 1s However at low shear rates the vis-cosity was observed to vary by up to an order of magnitude from the lowest to thehighest Mw Therefore the Mw of the PSE cannot only be used to tailor the deviceproperties but to simultaneously adjust the ink formulation for a given printingprocess without altering the chemical composition of the LECs
146 G Hernandez-Sosa et al
Fig 65 Current-voltage(top) and luminance-voltage(bottom) characteristics of thedifferent LECs shown inFig 64 prepared withdifferent PMMA Mw Themeasurement was conductedat room temperature and thevoltage was swept at a rate of01 Vs Inset photograph ofa LEC (4 6 mm) inoperation prepared with aPMMA Mw of 350000Reprinted from Ref [23]Copyright 2013 withpermission from Elsevier
Fig 66 Viscosity as afunction of shear rate for thesolutions prepared withdifferent PMMA Mw
Reprinted from Ref [23]Copyright 2013 withpermission from Elsevier
6 Morphology Engineering and Industrial Relevant Device hellip 147
624 Effect of the PSE Monomer Ratio on the FilmMorphology
The properties of polymers can be systematically tailored by the composition oftheir monomer unit In addition combining more than one of these chemical entitiesin a polymer chain (ie copolymerization) gives rise to new materials that combinethe properties of its components These intermediate properties are determined bythe ratio between the different monomers and the arrangement of the monomer units(ie block copolymerization vs random polymerization)
In this section we discuss the use of the copolymer poly(lactic-co-glycolic acid)(PLGA) as the base of a PSE in LECs comprising SY as a semiconductor [22]PLGA is composed of lactic acid and glycolic acid monomer units Consequentlyits mechanical and chemical properties can be tuned by varying the ratio betweenthese two components [31] Furthermore PLGA is a biocompatible andbiodegradable polymer commonly used in medical applications and clinical trials asdrug delivery agent or tissue scaffold [32 33] To date the use of PLGA inelectronics has not been thoroughly investigated However in the last several yearsthere has been an abundance of motivations to investigate the use of biocompatibleand biodegradable materials for electronic applications [34] On the one hand theuse of such materials helps to solve the growing electronic-waste problem On theother hand it enables medical applications such as bioresorbable electronicimplants Nevertheless the realization of sustainable biodegradablebiocompatiblelight-emitting devices at a large scale will require the understanding of the materialproperties the printing process and their relations to the device performance
In our recent work with PLGA we used electrochemical impedance spectroscopyto determine the ionic conductivity of SYPLGA[TBA][BF4] blends as shown inFig 67 We studied the three different lacticglycolic monomer weight ratios(8515 7525 6535 wt) with mass contents ranging from 0 to 142 wt relative to
Fig 67 Ionic conductivityof the device active layerdetermined by impedancespectroscopy and itsdependence on monomer ratioin the PLGA given in weightpercent The SY [TBA][BF4]ratio is always 102Reprinted from Ref [22]Copyright 2016 AmericanChemical Society
148 G Hernandez-Sosa et al
the SY plus salt mass It was observed that the ionic conductivity increases for allcases after the addition of PLGA The largest observed ionic conductivity resultedfrom using a PLGA with a 6535 lacticglycolic to monomer ratio at a 14 wt(r = 2 10minus9 Scm) This corresponded to 3 orders of magnitude improvementcompared to the pristine SY[TBA][BF4] blend which exhibited an ionic conduc-tivity of 3 10minus12 Scm Larger lactic monomer ratios of PLGA yielded a lesspronounced increase of the ionic conductivity reaching values of 10minus10 Scm
The improvement of the ionic conductivity suggests that the presence of PLGApromotes the movement of ions within the film This is demonstrated by theobserved decrease in operational voltages which were extracted from the corre-sponding luminancendashcurrentndashvoltage characteristicsmdashFig 68 In LECs the ionicconductivity affects the time that the ions need to accumulate in the regions at thecontacts enabling doping and charge injection into the semiconductorConsequently the ionic conductivity has an impact on the Von of the device whichwe defined as the voltage at which a luminance value of 1 cdm2 is measured InFig 68 it is observed that in all three cases the addition of 16 wt of PLGA isenough to decrease the Von by 2 V with respect to the reference devices withoutthe copolymer Interestingly upon increasing the amount of PLGA the behavior ofthe Von differs depending on the lacticglycolic monomer ratio In particular afterthe initial Von decrease the devices with the monomer ratio 6535 returns to itsoriginal value at a PLGA mass fraction of 11 wt even though the ionic con-ductivity steadily increases In contrast the other two investigated monomer ratiosshow a direct correlation between the increase in ionic conductivity and thedecrease in the Von when adding more PLGA This apparent contradiction wasclarified by the investigation of the film morphology
Figure 69 shows fluorescent microscopy (FLM) and AFM images of the LECactive layer morphology It presents measurements performed on samples withdifferent PLGA monomer ratios and mass fractions PLGA and [TBA][BF4] do not
Fig 68 Turn-on voltage as afunction of monomer ratioand the PLGA content givenin wt The SY[TBA][BF4]weight ratio is 102Reprinted from Ref [22]Copyright 2016 AmericanChemical Society
6 Morphology Engineering and Industrial Relevant Device hellip 149
fluoresce in the region of the band-pass filter (515ndash555 nm) of the fluorescencemicroscope therefore the bright and dark areas in the FLM images represent SYrich and SY poor phases respectively For the reference sample (SY[TBA][BF4])the SY emission was shown to be inhomogeneous throughout the image suggestingthe presence of [TBA][BF4] agglomerates A qualitatively more homogeneous lightemission is achieved when adding 16 wt PLGA suggesting that PLGA promotesa better distribution of [TBA][BF4]in the active layer This observation is inagreement with the higher ionic mobility and lower turn-on voltages shown inFigs 67 and 68 An overall phase separation between SY and PLGA withincreasing PLGA content can be observed in both the FML and AFM data Thistendency is more pronounced upon increasing the glycolic monomer ratio A largerglycolic monomer ratio in the PLGA polymer has been reported to produce a morehydrophilic polymer chain [35] which leads to worse intermixing with thehydrophobic SY as observed in Fig 69 Despite the larger ionic conductivitythe 6535 sample exhibited an increasing Von with increasing PLGA content On thebasis of the FLM images we conclude that the higher ionic conductivity can onlybe exhibited in the PLGA rich phases and therefore the observed phase separationdoes not favor electrochemical doping in the emitting material limiting deviceperformance
Fig 69 Fluorescent and atomic force micrographs demonstrating an increasing phase separationwith PLGA content and the respective glycolic acid monomer ratio Reprinted from Ref [22]Copyright 2016 American Chemical Society
150 G Hernandez-Sosa et al
The results highlighted in this example underline the importance of the micro-morphology for the device performance It also shows how varying the monomerratio of the PSE has a direct effect on the morphology ionic conductivity andconsequently the device performance This approach can be potentially used as auseful tool in the ink formulation of printed LECs In particular the study ofmaterials like PLGA not only addresses the study of an alternative device com-ponent but also provides insights into other aspects like sustainability or biocom-patibility of LECs
63 LEC Fabrication by Scalable Methods
631 Introduction
Since the original invention of the printing press graphical printing technologieshave tremendously evolved enabling the low-cost and high-throughput productionof newspapers magazines etc on a daily basis In the last decade one of theobjectives of the organic electronics field was the use of these technologies tofabricate optoelectronic devices in a cost efficient manner Therefore there is acontinuous need for research on solution processed electronics at a laboratoryscale to move away from the standard spin coating method to study the materialndashprocessndashdevice performance relations that govern printing and coating techniques
In contrast to graphical printing patterns optoelectronic devices have inherentlystricter requirements in terms of film thickness micromorphology and even theneed for well-defined multilayers in the nanometer thickness regime to achievefunctional devices Therefore the formulation of functional inks cannot be arbi-trarily changed to tailor their viscoelastic properties or adjust their drying times inorder to achieve better printability A change in the solvent or inclusion of additivescan greatly influence film morphology and consequently important properties of thefilm such as crystallinity or carrier mobility
A good understanding of the interplay between the material properties and thespecific printing technique is required to produce a functional ink formulation thatyields films suitable for device fabrication An ink formulation must be tailoredtowards the desired printing technique However the properties of the carrier fluidmight be restricted by intrinsic material properties such as molecular weight whichinfluences the viscosity or chemical structure which could affect wetting proper-ties The choice of printing technique depends mainly on the requirements of thedevice component andor the final application (eg necessary thickness require-ment for large-area coating 2-D structuring etc) Due to their working principlesthe different printing techniques will present advantages and disadvantages in termsof printing speed film thickness range or lateral resolution as well as some degreeof freedom to optimize the printing results For example gravure-cell volume forgravure printing screen mesh size for screen printing or drop volume for inkjet
6 Morphology Engineering and Industrial Relevant Device hellip 151
printing just to mention a few Figure 610 shows an image of a roll-to-roll printingpress used for the fabrication of electronics in which printing modules for differentprinting techniques can be combined
As previously discussed LECs possess inherent processing advantages overOLEDs which would favor their production by printing techniques The simplerdevice stack reduces the risk of dissolution and intermixing of the previouslyprinted layers The independence of device function on the electrode work functionallows for the fabrication of devices with air stable materials Additionally thelarger tolerance to film thickness variations lowers the requirements for film qualityand thickness control [17] In the previous section we presented two examples ofhow the molecular structure of the PSE has an impact on the overall properties ofthe LEC active layer We also indicated that a compromise between materialsproperties and processability is necessary to ultimately have a printed functionaldevice In this section we will discuss some methods and approaches that have beenused for the fabrication of LECs by industrially relevant techniques such as gravureand inkjet printing as well as slot die and spray coating
632 Gravure Printing
Gravure printing commonly used for the fabrication of magazines and packagingis a promising option for the high-throughput fabrication of optoelectronic devices
Fig 610 Image of a printing press used for printed electronics Press photo InnovationLabCopyright 2016 InnovationLab GmbH
152 G Hernandez-Sosa et al
Since most of its components are metallic it is suitable for and resistant against awide range of solvents At optimized conditions it can reach a printingspeed gt 10 ms Additionally it can achieve a lateral resolution on the micrometerscale This characteristic makes gravure printing one of the most promising printingtechniques for the large scale production of electronics [37] Figure 611 depicts thegravure-printing process In the first stage (a) the cells of the engraved gravurecylinder are filled with the functional ink In the ideal case the fluid should have aviscosity between 10 and 100 mPa s [18 37] During the second step a blade isused to remove excess ink (b) In the following step (c) the ink loaded cylinder isbrought in contact with the substrate in order to transfer the ink pattern This steprequires good wetting properties between the fluid and the substrate In the finalstage (d) two competing processes take place film drying and surface demodu-lation and leveling Films suitable for device fabrication will be produced only if thedrying time is longer than the leveling time [38]
6321 Gravure-Printed Polymer-Based LECs
In our previous work we have demonstrated the fabrication of gravure-printedactive layers and the fabrication of polymer-based LECs under ambient conditionsFigure 612 displays the viscosity of the investigated polymer blends consisting ofSY as semiconductor and the blend PMMA[TBA][BF4] as the PSE We investi-gated the influence of the blend composition the Mw of PMMA and the total solidconcentration on the printing outcome It was shown that by altering the blendcomposition we obtained a less pronounced shear-thinning of the viscosity Thiswas found to be beneficial for minimizing the viscous fingering pattern in the
Fig 611 Scheme of the gravure-printing process Reprinted from Ref [36] Copyright 2014Wiley
6 Morphology Engineering and Industrial Relevant Device hellip 153
printed outcome see Fig 612b and d The best material ratio was found to be1125025 for SYPMMA[TBA][BF4] This blend composition exhibited thehighest performance however also the shortest lifetime A second experimentconsisted of printing ink formulations at different solid concentrations (15 and10 gL) and using different PMMA Mw It was shown that the 10 gL formulationsgenerally showed more homogeneous layers due to a more Newtonian behavior ofthe viscosity The results of this experiment demonstrate that a polymer-based LECsystem offers the possibility of making ink formulations tailored to a given printingtechnique without changing its chemical composition
An image of the printed device under operation is shown in Fig 613 Theprinted device with the optimized formulation achieved maximum luminance val-ues of 2000 cdm2 and a Von of 65 V It also exhibited a shelf lifetime of sixmonths which would be sufficient for low-end applications like packaging oradvertisement
Fig 612 Shear-rate viscosity dependency for four different LEC ink formulations containingdifferent SYPMMA [TBA][BF4] ratios at a concentration of 10 gL (top left) Correspondingcontrast-enhanced photographs of gravure-printed films (top right) Shear-rate viscosity depen-dency of LEC ink formulations using four different Mw and two different solid concentrations(bottom left) SYPMMA [TBA][BF4] ratios were set constant (1125025) Correspondingcontrast-enhanced photographs of the gravure-printed films (bottom right) The inset shows thefast Fourier transform (FFT) of the corresponding image Printing field size is 225 cm2 Reprintedfrom Ref [36] Copyright 2014 Wiley
154 G Hernandez-Sosa et al
6322 Gravure-Printed Small Molecule-Based LECs
Small molecule emitters are one of the most promising approaches for the fabri-cation of high-performance light-emitting devices (see Chaps 9 and 12 for moredetails) However due to their low molecular weight they have a low viscosity insolution Such low viscosity hinders the processability using printing techniqueslike gravure printing therefore requiring a polymer matrix to enable printability[39] In our previous work we have demonstrated this by using an inert polymer inthe active layers of an OLED Here we opted to combine a small molecule emitter(SMG) with PMMA[TBA][BF4] for the fabrication of an LEC while simultane-ously enabling the use of gravure printing as a fabrication tool Figure 614 (left)shows the voltage-luminance characteristics of the reference LECs comprisingSMGPMMA[TBA][BF4] at a 10102 ratio deposited from anisole This spin-coated references exhibit Von 5 V and a maximum luminance 8000 cdm2Figure 614 (right) show the printing outcome of this ink formulation when varyingthe gravure-cell geometry parameters in the printing form instead of altering the ink
Fig 613 Current densityndashvoltagendashluminance characteristics of a gravure-printed LEC InsetPhotograph of a flexible LEC in operation Reprinted from Ref [36] Copyright 2014 Wiley
Fig 614 Luminance-voltage characteristics of a SMGPMMATBABF4 reference sample (left)Photographs of the printed fields under UV illumination (right) The size of the fields is 225 cm2
6 Morphology Engineering and Industrial Relevant Device hellip 155
formulation The images were taken under UV illumination A more homogeneouslayer is achieved when using a gravure form with 100 lines per cm (lncm)compared to lower line densities The change in the tone value from 90 to 100resulted in a change in film thickness from 70 to 90 nm respectively
633 Inkjet Printing
In the last years inkjet printing has gained a prominent place in the field of printedelectronics Because of its high printing resolution (20 microm) and more impor-tantly its flexibility due to the digital design of the printing pattern it can beutilized through the whole production chain from the ink optimization processdevelopment prototyping and upscaling without the need of refabricating printingforms [40]
One of the first reports of inkjet printed LECs was made by Mauthner et al [41]They reported the fabrication of LECs with a planar configuration using evaporatedgold electrodes The active layer was composed of MEH-PPV PEO andLiCF3SO3 all deposited from aqueous solution Two different molecular weights ofPEO were used in order to optimize the drop formation during the ink ejectionprocess Figure 615 demonstrates how the use of the low molecular weight PEO(30000 gmol) led to the elimination of the bead-on-a-string effect present whenusing the PEO with 100000 gmol Mw It was further shown that the optimizedprinted devices showed a Von 3 V and the typical red emission of MEH-PPV
An interesting approach to the fabrication of inkjet printed LEC devices waspresented by Lindh et al [42] In this work the authors opted for a bilayer con-figuration PSEsemiconductor As schematically presented in Fig 616 the processconsisted of printing first the PSE (polyethylene glycol (PEG) mixed withKCF3SO3) as a dotted pattern on to a transparent electrode Subsequently theluminescent layer and the complementary electrode were deposited on top Withisolated islands of PSE device operation relied on the drift of ions into the semi-conductor with applied bias causing light emission to originate from regions inclose proximity to the PSE-semiconductor interface The devices were reported toachieve a power conversion efficacy of 043 lmW and were used to display staticmessages with a pixel density of 170 pixels per inch (PPI) over several hours as canbe seen in Fig 617
634 Slot-Die Coating
Slot-die coating is a film deposition technique which was commonly utilized forthe production of photographic film Although it is not a printing technology due tothe lack of 2-D structuring slot-die coating is a promising technology for large-areathin film devices due to the homogeneity of the deposited layers Indeed it allows
156 G Hernandez-Sosa et al
to prepare film at speeds gt10 ms and has been used for the deposition of organicsolar cells with promising results [17 18 43]
The most relevant study on slot-die coated LECs to date was performed by thegroups of L Edman and F C Krebs [17] The entire process was carried out underambient conditions as can be seen in Fig 618 They demonstrated a multilayerprocess on flexible substrates using a slot-die roll-coating equipment operatedat 06 mmin The devices displayed uniform light emission and a fault-tolerantactive layer with a thickness gt1 lm An example of such devices is presented inFig 619
Fig 615 Drop formation of the MEH-PPV dispersion and of the blend systems (left)Microscope image schematic and electroluminescence of a planar inkjet printed LEC (right)Reprinted from Ref [41] Copyright 2008 Elsevier
Fig 616 Schematics of the parallel inkjet printing of electrolyte-ink droplets on top of atransparent bottom electrode (left) and one lattice cell in a completed bilayer sandwich device(center) A photograph of a portion of an inkjetted lattice of dry electrolyte droplets deposited ontop of a transparent ITO electrode (right) Reprinted from Ref [42] Copyright 2014 Wiley
6 Morphology Engineering and Industrial Relevant Device hellip 157
The ink formulation utilized by Edman and coworkers was based on a blendcomposed of SY and a PSE based on PEOKCF3SO3 dissolved in cyclohexane at aratio of 1135025 A maximum efficacy of 06 cdA at a corresponding brightnessof 50 cdm2 was reached for the champion device This relatively low performancecompared to state of the art was caused by the high relative content of PSE in theactive layer which was shown to be detrimental for device operation [44]However such a high PSE ratio was necessary to ensure optimum coating of layers
Fig 617 Optoelectronic characterization of a bilayer LEC with a filled lattice pattern and a drydroplet diameter and pitch of 50 and 90 lm respectively The device was driven with a constantcurrent of 1 mA and the active area was 113 mm2 The inset shows the homogenous emissionfrom such a bilayer LEC with a droplet diameter and pitch of 60 and 120 lm respectivelyReprinted from Ref [42] Copyright 2014 Wiley
Fig 618 Photograph of the slot-die roll coater during the deposition of the active layer (left)Close-up photograph of the slot-die head during coating of an active layer stripe (right) Adaptedby Ref [17] Copyright 2012 Macmillan Publishers Ltd
158 G Hernandez-Sosa et al
Nevertheless further optimization of the process and materials is expected to leadto the state-of-the art efficiencies attained in laboratory experiments but with thisindustrially relevant technique
635 Spray Coating
Spray coating is a versatile technique that is suitable for a large variety of materialsand solvents It is capable of accurate thickness control [46] with the possibility of2-D structuring by means of a shadow mask It has been proven effective inapplication fields such as paintings and protective coatings and has also been usedfor the fabrication of organic electronic devices [47 48] In general the desiredmorphology of the deposited film can easily be tuned by varying the process andmaterial parameters such as solid concentration substrate temperature carrier-gasflow rate or nozzle to substrate distance [46]
Sandstroumlm et al used spray coating to fabricate LECs with an emitting areaof 200 cm2 using the ldquospray-sinteringrdquo method [45] This method was developedto produce an active layer with micro-particle morphology rendering the devicetolerant to defects and dust particles Specifically the sprayed droplets should startevaporating before reaching the substrate in order to prevent the formation of a wetfilm once the droplets land on the substrate The droplets should then ldquosinterrdquotogether during the rest of the drying process forming a particle network capable toencapsulate dust particles or seal pinholes An example of such an active layer ispresented in Fig 620 The image shows circular-shaped features with diametersof 20 lm
The authors also demonstrated the fabrication of multilayer structures andmulticolor emission The versatility of the spray coating technique allowed them tofabricate devices onto arbitrary shapes such as a fork which would not be possibleto do with any other sheet or roll based process (see Fig 621) In terms of
Fig 619 Photograph of a slot-die coated LEC illustrating the bidirectional light emission andthe device conformability (left) Light emission from a semitransparent slot-die coated LECfollowing gt6 months storage in a glove box (right) The devices were driven at V = 7 V Adaptedby Ref [17] Copyright 2012 Macmillan Publishers Ltd
6 Morphology Engineering and Industrial Relevant Device hellip 159
Fig 620 Optical microscopy images of the first monolayer deposited during spray-sinteringwith clearly distinguishable dry particle features Reprinted from Ref [45] Copyright 2014 Wiley
Fig 621 Ambient spray-sintering of entire LEC structures and the inclusion of light-emissionfunction onto flat and complex-shaped metallic surfaces Top Schematic depicting the sequentialspray-sintering of a bottom electrode an active layer and a top electrode Bottom left Lightemission from an all-spray-sintered LEC driven at V = 5 V Bottom middle Light emission froman all-spray-sintered LEC using an Al plate coated with a layer of PEDOTPSS as the combinedsubstrate and anode Bottom right A light-emitting fork as realized by spray-sintering astainless-steel fork with an active layer and a top cathode Reprinted from Ref [45] Copyright2014 Wiley
160 G Hernandez-Sosa et al
performance the spray sintered LEC showed an identical performance to thespin-cast reference reaching an efficacy of 58 cdA at 225 cdm2 and 60 cdA at233 cdm2 respectively
64 Conclusion
In this chapter we have reviewed several approaches to control morphology of PSEbased LECs as well as examples of LECs processing by industrially scalableprinting and coating techniques In the first part of the chapter it was shown that thePSE molecular structure is not only relevant for its function in the device (ie ionsolvating material) but also for its critical role in the morphology and rheologicalproperties of the active layer blend A change to its molecular weight or monomerratio was shown to affect phase separation and consequently device performanceAdditionally these properties were also shown to affect the printability of the layer
In the second section several examples of the fabrication of LECs by industrialrelevant techniques were presented The results of gravure and inkjet printing aswell as on slot-die and spray coating demonstrated that the properties of the PSEwas decisive for an optimum ink formulation that enables device fabrication Thehighlighted results demonstrate that the steps taken to adjust a functional ink for-mulation will inevitably result in a compromise between performance and print-ability In the future however the resulting drawbacks could be minimized if thespecific characteristics of the printing process are taken into account during theearly stages of the material development The device performances achieved inthe presented examples serve as a strong motivation to continue research in the fieldand predicts a bright future for LEC-based technological applications produced byindustrial means
Acknowledgements This work was partially supported by the Federal Ministry for Educationand Research grant numbers 03X5526 and 13N11903 The authors are grateful to M HamburgerS Stolz R Eckstein U Lemmer for fruitful discussions
References
1 C Tang S VanSlyke Appl Phys Lett 51 913 (1987)2 G Chansin K Ghaffarzadeh H Zervos OLED Display Forecasts 2016ndash2026 The Rise of
Plastic and Flexible Displays (2016)3 B Geffroy P Le Roy C Prat Pol Int 55 572 (2006)4 S Reineke F Lindner G Schwartz N Seidler K Walzer B Luumlssem K Leo Nature
459 234 (2009)5 R Das P Harrop Printed Organic amp Flexible Electronics Forecasts Players amp
Opportunities 2016ndash2026 (2016)6 P Matyba H Yamaguchi G Eda M Chhowalla L Edman ND Robinson ACS Nano
4 637 (2010)
6 Morphology Engineering and Industrial Relevant Device hellip 161
7 L Edman Electrochim Acta 50 3878 (2005)8 P Matyba K Maturova M Kemerink ND Robinson L Edman Nature Mater 8 672
(2009)9 SB Meier D Tordera A Pertegaacutes C Roldaacuten-Carmona E Ortiacute HJ Bolink Mater Today
17 217 (2014)10 A Sandstroumlm L Edman Energy Tech 3 329 (2015)11 SB Meier D Hartmann D Tordera HJ Bolink A Winnacker W Sarfert Phys Chem
Chem Phys 14 10886 (2012)12 RD Costa E Ortiacute HJ Bolink Pure Appl Chem 83 2115 (2011)13 Q Pei G Yu C Zhang Y Yang AJ Heeger Science 269 1086 (1995)14 SB Meier S van Reenen B Lefevre D Hartmann HJ Bolink A Winnacker W Sarfert
M Kemerink Adv Funct Mater 23 3531 (2013)15 F Huang H Wu Y Cao Chem Soc Rev 39 2500 (2010)16 S Stolz M Petzoldt N Kotadiya T Roumldlmeier R Eckstein J Freudenberg UH Bunz
U Lemmer E Mankel M Hamburger et al J Mater Chem C 4 11150 (2016)17 A Sandstroumlm HF Dam FC Krebs L Edman Nature Commun 3 1002 (2012)18 H Kipphan Handbook of Print Media Technologies and Production Methods (Springer
Science amp Business Media 2001)19 S Tang L Edman J Phys Chem Lett 1 2727 (2010)20 D Tordera S Meier M Lenes RD Costa E Ortiacute W Sarfert HJ Bolink Adv Mater
24 897 (2012)21 J Mindemark L Edman J Mater Chem C 4 420 (2016)22 J Zimmermann N Juumlrgensen AJ Morfa B Wang S Tekoglu G Hernandez-Sosa ACS
Sust Chem Eng (2016)23 G Hernandez-Sosa R Eckstein S Tekoglu T Becker F Mathies U Lemmer N Mechau
Org Elec 14 2223 (2013)24 P Matyba MR Andersson L Edman Org Elec 9 699 (2008)25 L Sardone CC Williams HL Anderson G Marletta F Cacialli P Samori Adv Funct
Mater 17 927 (2007)26 FP Wenzl P Pachler C Suess A Haase EJ List P Poelt D Somitsch P Knoll
U Scherf G Leising Adv Funct Mater 14 441 (2004)27 J Chappell DG Lidzey PC Jukes AM Higgins RL Thompson SOrsquoConnor I Grizzi
R Fletcher J OrsquoBrien M Geoghegan et al Nat Mater 2 616 (2003)28 E Moons J Phys Condensed Matter 14 12235 (2002)29 RD Costa A Pertegaacutes E Ortiacute HJ Bolink Chem Mater 22 1288 (2010)30 Y Shao GC Bazan AJ Heeger Adv Mater 19 365 (2007)31 JM Anderson MS Shive Adv Drug Deliv Rev 64 72 (2012)32 B Dhandayuthapani Y Yoshida T Maekawa DS Kumar Int J Polit Sci 2011 (2011)33 HK Makadia SJ Siegel Polymers 3 1377 (2011)34 M Irimia-Vladu ED Glowacki G Voss S Bauer NS Sariciftci Mater Today 15 340
(2012)35 DA Norris N Puri ME Labib PJ Sinko J Control Rel 59 173 (1999)36 G Hernandez-Sosa S Tekoglu S Stolz R Eckstein C Teusch J Trapp U Lemmer
M Hamburger N Mechau Adv Mater 26 3235 (2014)37 RR Soslashndergaard M Houmlsel FC Krebs J Polym Sci B Polym Phys 51 16 (2013)38 G Hernandez-Sosa N Bornemann I Ringle M Agari E Doumlrsam N Mechau U Lemmer
Adv Funct Mater 23 3164 (2013)39 S Tekoglu G Hernandez-Sosa E Kluge U Lemmer N Mechau Org Elect 14 3493
(2013)40 L Wu Z Dong F Li H Zhou Y Song Adv Opt Mater (2016)41 G Mauthner K Landfester A Koumlck H Bruumlckl M Kast C Stepper EJ List Org Elect
9 164 (2008)42 EM Lindh A Sandstroumlm L Edman Small 10 4148 (2014)
162 G Hernandez-Sosa et al
43 FC Krebs N Espinosa M Houmlsel RR Soslashndergaard M Joslashrgensen Adv Mater 26 29(2014)
44 J Fang P Matyba L Edman Adv Funct Mater 19 2671 (2009)45 A Sandstroumlm A Asadpoordarvish J Enevold L Edman Adv Mater (2014)46 A Abdellah B Fabel P Lugli G Scarpa Org Elect 11 1031 (2010)47 A Falco A Zaidi P Lugli A Abdellah Org Elect 23 186 (2015)48 A Morfa T Roumldlmeier N Juumlrgensen S Stolz G Hernandez-Sosa Cellulose 23 3809
(2016)
6 Morphology Engineering and Industrial Relevant Device hellip 163
Part IIITraditional and New
Electroluminescent Materials
Chapter 7Development of CyclometallatedIridium(III) Complexes for Light-EmittingElectrochemical Cells
Catherine E Housecroft and Edwin C Constable
Abstract This chapter gives an overview of the development of cyclometallatediridium(III) complexes for application in light-emitting electrochemical cells(LECs) and highlights the ligand-design strategies employed to enhance devicestability operating efficiency and (critically for LECs in which the ion mobilitiesare typically low) turn-on times Typical iridium-containing ionic transition metalcomplexes (Ir-iTMCs) belong to the family of [Ir(C^N)2(N^N)]
+ complexes inwhich H(C^N) is a cyclometallating ligand and N^N is a diimine or relatedchelating ligand The partitioning of IrC^N versus N^N character in the HOMOand LUMO of a [Ir(C^N)2(N^N)]
+ complex respectively lends itself to a ligandfunctionalization-driven method of varying the band-gap allowing emission-colourtuning An important development in addressing device stability has been thedesign of ligands that can protect the iridium(III) centre through intra-cationp-stacking interactions and progress in this area is discussed in detail The need fordeep-blue emitters has been addressed by introducing fluoro-substituents into thecyclometallating domain however this can lead to lower stability of the LECs andalternative means of shifting the emission into the blue are briefly discussed (seeChap 8 for an in-depth discussion) Finally we discuss how a move away fromsingly charged cationic Ir-iTMCs can be used to shorten the turn-on times of LECs
Keywords Light-emitting electrochemical cell Iridium Ligand functionaliza-tion Ligand modification Device stability Device turn-on time
71 Introduction
Light-emitting electrochemical cells (LECs) are light-emitting devices in which theactive material is a charged species In the prototype devices luminescent polymerscontaining ionic salts were used [1] The first LEC incorporating an ionic transition
CE Housecroft (amp) EC ConstableDepartment of Chemistry University of Basel Spitalstrasse 51 4056 Basel Switzerlande-mail catherinehousecroftunibasch
copy Springer International Publishing AG 2017RD Costa (ed) Light-Emitting Electrochemical CellsDOI 101007978-3-319-58613-7_7
167
metal complex (iTMC) as the active compound was reported in 1996 [2] andcontained a [Ru(bpy)3]
2+-based emitter (bpy = 22prime-bipyridine) Two problemslimit the development of Ru-iTMCs (i) the difficulty of colour tuning the emissionaway from the characteristic orange-red of [Ru(bpy)3]
2+ and (ii) the low stability ofthe active materials under LEC device conditions By replacing ruthenium(II)complexes with those of iridium (from second to third row d-block) not only isthe stability enhanced but it also becomes easier to tune the emission energy bysystematic functionalization of the ligands The archetype iridium(III) tris-chelatethat replaces [Ru(bpy)3]
2+ is [Ir(ppy)2(bpy)]+ (Fig 71) where Hppy = 2-
phenylpyridine and [ppy]ndash is a cyclometallating CN-donor ligand Figure 72gives a schematic representation of a double-layer LEC device which includes ahole injection layer (poly(34-ethylenedioxythiophene)-poly(styrenesulfonate) =
Fig 71 The structure [17] and a schematic representation of the archetype complex[Ir(ppy)2(bpy)]
+ The trans-arrangement of the N-donors of the two [ppy]ndash ligands is typical of[Ir(C^N)2(N^N)]
+ complexes
Fig 72 A working LEC and a schematic representation of a double-layer LEC (ITO = indiumtin oxide PEDOTPSS = poly(34-ethylenedioxythiophene)polystyrenesulfonate IL = ionic liq-uid) Metals such as Ag or Au may replace Al as the cathode (Photograph Dr Collin D MorrisUniversity of Basel)
168 CE Housecroft and EC Constable
PEDOTPSS) In the simplest device architecture the hole injection layer is absentThe active material is an iTMC and in this chapter we focus on Ir-iTMCs [3ndash14]other chapters review other active materials including polymers small moleculescopper(I) complexes quantum dots and nanoparticles (see Chaps 9 10 11 12 and13) The electroluminophore is usually blended with an ionic liquid (IL) such as1-butyl-3-methylimidazolium hexafluoridophosphate ([BMIM][PF6]) use of ILimproves the ionic conductivity of the active layer [15 16] Some of the challengesthat face the LEC research community are (i) the shortening of the turn-on time ofthe LEC device (defined either as the time for the device to attain maximumluminance or the time to reach a defined luminance) (ii) increasing the devicelifetime (the lifetime t12 is defined as the time for the luminance (or brightness) todecay from its maximum value Bmax to half this value Bmax2 ndash see Chap 1 formore details) and (iii) achieving the holy-grail of white light emitters If singlecomponent white light emitters cannot be effectively developed then there is a needfor deep-blue emitters to complement the widely available palette of orangeemitters in order to achieve a two-component white
In this chapter we pay particular attention to the structural features of[Ir(C^N)2(N^N)]
+ complexes (C^N denotes a CN-donor ligand typically but notalways of the anionic cyclometallated type typified by [ppy]ndash and N^N is a neutralbidentate NN-donor typified by 22prime-bipyridine) that influence device stabilityturn-on time and t12 We also demonstrate how the choice of ligand affects theemission wavelength (colour tuning) A significant property of [Ir(C^N)2(N^N)]
+
complexes is that in the ground state the highest-occupied molecular orbital(HOMO) is localized on the Ir(C^N)2 domain whilst the lowest-unoccupiedmolecular orbital (LUMO) is localized on the N^N ligand (Fig 73) The discreteorbital composition of the [Ir(C^N)2(N^N)]
+ HOMOndashLUMO manifold readilylends itself to colour tuning of the emission wavelength by designing C^N ligandsin order to modify the HOMO energy and by functionalizing N^N ligands to
Fig 73 Representations of the HOMO (left) and LUMO (right) of the archetype Ir-iTMC [Ir(ppy)2(bpy)]
+ cation which are localized respectively on the Ir(ppy)2 and N^N domains(calculated at DFT level (B3LYP 6-31G) using Spartan 16 Wavefunction Inc)
7 Development of Cyclometallated Iridium(III) Complexes hellip 169
modify the LUMO energy level We shall see this strategy at work throughout thischapter Since the HOMO is located on the Ir(C^N)2 unit the IrIIIIrIV oxidationpotential (Eox) gives a measure of the HOMO level and the examples in Table 71illustrate tuning of the HOMO energy by functionalization of the bpy ligand in[Ir(ppy)2(N^N)][PF6] complexes In addition Table 1 in reference [10] gives auseful summary of photophysical and electrochemical properties of a range ofIr-iTMCs High-level theoretical studies of iTMCs in both the ground and excitedstates are pivotal to an understanding of experimental results and are invaluable indesigning new emissive materials [10 18ndash23]
72 Synthetic Approaches to [Ir(C^N)2(N^N)]+ Complexes
721 Use of [Ir2(C^N)4(l-Cl)2] Dimers
The most common synthetic route to [Ir(C^N)2(N^N)]+ cations is by formation of
the chlorido-bridged [Ir2(C^N)4(l-Cl)2] dimer (Fig 74 left) followed by treatmentwith the N^N ancillary ligand The dimer is typically prepared by reaction ofIrCl3xH2O with the cyclometallating ligand H(C^N) under reflux conditions(Scheme 71) following the strategy of Watts and coworkers [27] Although thismethodology is widely applied it may be inefficient when the cyclometallatingligand contains reactive substituents for example a sulfone [28] An alternativeroute is the reaction of [Ir2(cod)2(l-Cl)2] with H(C^N) (cod = 15-cyclooctadiene)This route has for example been used to prepare mixtures of heteroleptic dimers[Ir2(ppy)4ndashn(dfppy)n(l-Cl)2] (Hdfppy = 2-(24-difluorophenyl)pyridine) [29]
Salts of [Ir(C^N)2(N^N)]+ are conveniently prepared by treatment of
[Ir2(C^N)4(l-Cl)2] with the appropriate N^N ligand in methanol in a microwavereactor at 120 degC or by heating the reagents in a solvent such as methanol understandard reflux conditions In most cases the cyclometallated complexes are isolatedas hexafluoridophosphate salts by addition of [NH4][PF6] to the reaction mixture
Table 71 Variation in IrIIIIrIV oxidation potential (Eox) with N^N ligand functionalizationValues are referenced with respect to the FcFc+ couple
Complexa Eox (V) Solvent Reference
[Ir(ppy)2(bpy)][PF6] +084 DMF [24]
[Ir(ppy)2(6-Mebpy)][PF6] +089 DMF [24]
[Ir(ppy)2(66prime-Me2bpy)][PF6] +093 DMF [24]
[Ir(ppy)2(44prime-Ph2bpy)][PF6] +095 DMF [24]
[Ir(ppy)2(55prime-Ph2bpy)][PF6] +084 MeCN [25]
[Ir(ppy)2(44prime-tBu2bpy)][PF6] +091b MeCN [25]
[Ir(ppy)2(44prime-(Me2N)2bpy)][PF6] +072 MeCN [26]aLigand structures are given in Sect 73 bThe literature value was measured with respect to SCEthen adjusted to FcFc + (040 V vs SCE)
170 CE Housecroft and EC Constable
(Scheme 72) In a few studies the effect of the counter-ion on device performancehas been examined [32] with precipitation of the appropriate salt by the addition ofammonium or sodium salts of [BF4]
ndash [BPh4]ndash or [BIm4]
ndash (Scheme 72)Although these routes are used for the vast majority of syntheses of
[Ir(C^N)2(N^N)]+ luminophores traces of residual chloride counter-ion can have a
significant and detrimental effect on LEC device performance This has beendemonstrated with the prototypical complex [Ir(ppy)2(bpy)][PF6] where tightion-pairing between aromatic protons and chloride ion (established both in solutionand the solid state Fig 75) led to chloride ion being carried through the synthesisfrom [Ir2(ppy)4(l-Cl)2] to [Ir2(ppy)4(bpy)][PF6Cl] [33] In the 1H NMR spectrumof the product characteristic shifting of the bpy H3 proton (which interacts with theClndash ion Fig 75) can be used to quantify even small traces of chloride ion
Fig 74 Left Crystallographically determined structure of [Ir2(ppy)4(l-Cl)2] (CSD refcodeRINXES [30]) as an example of an [Ir2(C^N)4(l-Cl)2] dimer Right Crystallographicallydetermined structure of [Ir(ptrz)2(NCMe)2]
+ (CSD refcode QOMGAC [31]) the tetrazole-basedcyclometallated complex cation shown in Scheme 74
Scheme 71 Preparation of [Ir2(ppy)4(l-Cl)2] as an example of the most common route used forthe formation of the critical [Ir2(C^N)4(l-Cl)2] precursors
7 Development of Cyclometallated Iridium(III) Complexes hellip 171
722 Solvento Complexes
The use of solvento-iridium(III) precursors is a strategy developed to overcome theunfavourable effects of chloride-impurities described above Watts and coworkers[34] first introduced this approach by treating [Ir2(ppy)4(l-Cl)2] dimers with solublesilver salts which removed chloride ion by precipitation of AgCl The method(Scheme 73) has subsequently been used to isolate luminophores in LECs whichexhibit exceptional stabilities [35] A solvento approach in which the intermediateinvolves coordinated MeCN has been used for the preparation of tetrazole-based
Scheme 72 Conversion of [Ir2(ppy)4(l-Cl)2] to [Ir2(ppy)4(bpy)][X] as a typical example for thepreparation of salts of [Ir(C^N)2(N^N)]
+ complexes
Fig 75 Ion-pairing between[Ir(ppy)2(bpy)]
+ and Clndash in thesolid-state structure of 2[Ir(ppy)2(bpy)][Cl]2CH2Cl2[H3O][Cl] (CSD refcodeRISMOX [33])
Scheme 73 Solvento-intermediate approach to chloride-free salts of [Ir(C^N)2(N^N)]+ illus-
trated by the preparation of [Ir(ppy)2(bpy)][PF6]
172 CE Housecroft and EC Constable
cyclometallated complexes (Scheme 74) In this case attempts to isolate the cor-responding chlorido dimer by the direct reaction between IrCl3xH2O with theN-substituted 5-phenyltetrazole failed The solvento-complex shown in Scheme 74was structurally characterized (Fig 74 right) [31]
73 Development of Ligand Types in [Ir(C^N)2(N^N)]+
Emitters
The ligands referred to in this section are defined in Schemes 75 76 78 and 79
731 Archetype [Ir(ppy)2(bpy)]+ and [Ir(ppy)2(phen)]
+
Complexes
Detailed investigations of the photophysical behaviour of the archetype complexes[Ir(ppy)2(bpy)]
+ and [Ir(ppy)2(phen)]+ and their performances in LECs combined
with a theoretical understanding of the ground and excited state orbital composi-tions and energies have played an important role in defining targets for furtherligand development [17] The photoluminescence (PL) emission spectra of[Ir(ppy)2(bpy)]
+ and [Ir(ppy)2(phen)]+ in solution (kem
max = 585 and 530 nmrespectively) and in PMMA-based thin films (PMMA = poly(methyl methacrylate)show unstructured broad bands typical for metal-to-ligand charge transfer (MLCT)electronic transitions As has now been observed in many studies the photolumi-nescence quantum yields () increase on going from fluid solution to thin film thelatter being relevant for the LEC device although the observation of a photolu-minescence is not a guarantee that the same compound in a LEC device will exhibitmore efficient electroluminescence The increases from 14 to 66 for [Ir(ppy)2(bpy)]
+ and from 17 to 71 for [Ir(ppy)2(phen)]+ on moving from fluid
solution to PMMA film Electroluminescence (EL) spectra exhibit maxima at 590and 578 nm for [Ir(ppy)2(bpy)]
+ and [Ir(ppy)2(phen)]+ respectively Working with
structurally simple luminophores this benchmarking study showed that a LEC
Scheme 74 Solvento-intermediate approach to tetrazole-based cyclometallated complexes
7 Development of Cyclometallated Iridium(III) Complexes hellip 173
Scheme 75 Structures and abbreviations of N^N ligands in Sect 73
174 CE Housecroft and EC Constable
using [Ir(ppy)2(bpy)][PF6] exhibited high luminance (Bmax = 219 cdm2) and highstability (t12 = 668 h) whilst that with [Ir(ppy)2(phen)][PF6] performed less well(Bmax = 40 cdm2 t12 = 567 h) [17] Both LECs show long turn-on times (ton)indicating poor ion mobility but these can be shortened by the use of ionic liquidWith ratios of Ir-iTMCIL of 10 41 and 11 ton decreases from 702 to 72 to 07 hfor [Ir(ppy)2(bpy)][PF6] and from 160 to 64 to 009 h for [Ir(ppy)2(phen)][PF6]Both LECs show enhanced Bmax with the use of IL but the t12 are also severelyreduced Using [Ir(ppy)2(bpy)][PF6] enhanced luminances reduced values of tonand long t12 can also be achieved by the addition of [X][PF6] (X
+ = NH4+ K+ or
Li+) [36] Addition of Li[PF6] similarly boosts performance of LECs containing[Ir(24-Ph2py)2(bpy)][PF6] [37]
732 Functionalizing N^N Ligands in [Ir(ppy)2(bpy)]+
and [Ir(ppy)2(phen)]+ with Bulky Substituents
In this section we look at the effects of introducing substituents into the archetypeN^N ligands with a focus on sterically demanding substituents introduced in the33prime- 44prime- or 55prime-positions of bpy or 47-positions of phen Such functionalizationtypically leads to stable and efficient LEC devices The PL and EL spectra of[Ir(ppy)2(47-Ph2phen)][PF6] show similar maxima at 600 nm A LEC with[Ir(ppy)2(47-Ph2phen)][PF6] achieves a t12 = 65 h which at the time (2006) wasthe longest lived solid-state LEC device using an Ir-iTMC [38] The decay inluminance is accompanied by a decay in current density It was concluded thatthe long t12 of the LEC containing [Ir(ppy)2(47-Ph2phen)]
+ is largely due tothe hydrophobicity and steric hindrance of the peripheral phenyl rings in47-Ph2phen which suppress reactions with water and its degradation productsA systematic investigation of the photophysical properties of [Ir(ppy)2(bpy)]
+
Scheme 76 Structures and abbreviations of cyclometallating ligands in Sect 73
7 Development of Cyclometallated Iridium(III) Complexes hellip 175
[Ir(ppy)2(6-Mebpy)]+ [Ir(ppy)2(66prime-Me2bpy)]+ and [Ir(ppy)2(66prime-Me2-44prime-
Ph2bpy)]+ and of their performances in LECs has been complemented by a theo-
retical investigation [24] Values of t12 are significantly enhanced with increasedsubstitution and the PL of the Ir-iTMC in a film of the same composition as thatused in the LEC approximately doubles (34ndash74) on going from [Ir(ppy)2(bpy)]
+
to [Ir(ppy)2(66prime-Me2-44prime-Ph2bpy)]+ The increased substitution also has a benefi-
cial effect on the LEC stability which can be further improved by altering the ratioof Ir-iTMCIL (Table 72) Similarly in a series of [Ir(ppy)2(29-X2phen)]
+ lumi-nophores in which X = Me nBu or Ph increasing the size of the substituents leadsto improved luminance [39] A 55prime-substitution pattern in bpy is also advantageousSingle-layer LECs of configuration ITOIr-iTMCAu with an active component of[Ir(ppy)2(55prime-Ph2bpy)][PF6] exhibit high stability with values of t12 107 h [25]This surpasses the t12 of a similarly configured LEC containing[Ir(ppy)2(44prime-
tBu2bpy)][PF6] but the time to reach the maximum emission isfar longer for the LEC with [Ir(ppy)2(55prime-Ph2bpy)][PF6] (140 h) than[Ir(ppy)2(44prime-
tBu2bpy)][PF6] (2 h) Replacing the tBu substituents by stronglyelectron-donating NN-dimethylamino groups in [Ir(ppy)2(44prime-(Me2N)2bpy)][PF6]leads to a blue-green emitter the shift in PL emission wavelength arising fromdestabilization of the LUMO kem
max = 581 nm in [Ir(ppy)2(44prime-tBu2bpy)][PF6] and
520 nm with 491(sh) nm in [Ir(ppy)2(44prime-(Me2N)2bpy)][PF6] [26] Thesesingle-layer LECs emit light with an onset voltage as low as 25 V
The incorporation of 99-dihexylfluorenyl groups into the periphery of the N^Nligands 3-fluorphen and 38-(fluor)2phen (Scheme 75) and analogous N^N ligandsin which the n-hexyl chains are terminated by carbazole units gave a series ofligands designed to extend the p-conjugation in the Ir-iTMC Increased degree ofsubstitution of phen leads to increased excited state lifetimes (s) and the biasrequired to turn on and drive LECs of configuration ITOPEDOTPSSIr-iTMCAlincreased with the size and number of side-chains The LEC device stability were inthe range of t12 = 6ndash20 h under a 6 V bias [40 41] The N^N ligand diazaspiro(Scheme 75 and Fig 76) has been designed to incorporate both electron- and
Table 72 Effects of substituents and use of ionic liquid on LEC device t12 LECconfiguration = ITOPEDOTPSSIr-iTMCILAl and IL = [BMIM][PF6] Each complex wasused as a [PF6]
ndash salt Data from Ref [24]
Complex Ir-iTMCIL t12 (h)
[Ir(ppy)2(bpy)]+ 4 1 70
[Ir(ppy)2(bpy)]+ 1 1 78
[Ir(ppy)2(6-Mebpy)]+ 4 1 269
[Ir(ppy)2(6-Mebpy)]+ 1 1 28
[Ir(ppy)2(66prime-Me2bpy)]+ 4 1 223
[Ir(ppy)2(66prime-Me2bpy)]+ 1 1 41
[Ir(ppy)2(66prime-Me2-44prime-Ph2bpy)]+ 4 1 356
[Ir(ppy)2(66prime-Me2-44prime-Ph2bpy)]+ 1 1 95
176 CE Housecroft and EC Constable
hole-transporting domains Disappointingly the of [Ir(ppy)2(diazaspiro)][PF6] inthe solid-state is lt1 because the long s of 23 ms facilitates radiationless deac-tivation pathways Not unexpectedly the LEC device performance is poor [42]
Efficient orange LECs have been achieved using [Ir(ppy)2(imp)][PF6] (seeScheme 75 for the structure of imp) The solution PL of the complex is charac-terized by kem
max = 583 nm = 43 and an s of 910 ns A LEC with configurationof ITOPEDOTPSSIr-iTMCIL (4 1)Al showed a fast ton of 28 min to reachBmax of 684 cdm2 with only 45 s needed to reach a benchmark luminance of100 cdm2 Use of this ancillary ligand is very promising with the device losingonly 20 of its efficacy (Effmax = 65 cdA) during 850 h of operation [43]
733 Cyclometallating Ligands with Nitrogen-RichHeterocycles
The LEC literature encompasses many Ir-iTMCs with nitrogen-rich heterocycles ineither the C^N or N^N domains In this section we have chosen a number of simplesystems to illustrate the effects of changing the pyridine ring in Hppy by a pyrazoletriazole tetrazole or imidazole allowing comparisons to be made with [ppy]ndash
systems described aboveThe triplet state energies of the [ppy]ndash and bpy ligands are similar those of
[ppz]ndash and bpy are well separated and [Ir(ppz)2(N^N)][PF6] complexes exhibitemission maxima that are at higher energy (blue-shifted) with respect to the anal-ogous [Ir(ppy)2(N^N)][PF6] complexes This was demonstrated by Thompsonand coworkers [44] for neutral [Ir(ppz)3] versus [Ir(ppy)3] complexes and was
Fig 76 The solid-statestructure of [Ir(ppy)2(diazaspiro)]
+ (CSDrefcode EJIZUT [42]) whichincorporates both electron-and hole-transportingdomains in the N^N ligand
7 Development of Cyclometallated Iridium(III) Complexes hellip 177
later extended to cationic [Ir(ppz)2(N^N)]+ complexes A LEC containing
[Ir(ppz)2(44prime-tBu2bpy)][PF6] took 70 min to reach an external quantum effi-
ciency (EQE) of 69 under a 3 V bias and gave a peak power efficiency of25 lmW [45] We described earlier that the inclusion of bulky substituents in theN^N domain is beneficial Similarly LECs containing [Ir(ppz)2(47-Ph2phen)][PF6]or [Ir(ppz)2(3478-Me4phen)][PF6] achieve high luminances Their EL kem
max are574 and 537 nm respectively [46] which are blue-shifted with respect to thosewith the analogous [ppy]ndash-containing luminophores [38] Maximum luminancesof 5199 and 4751 cdm2 for LECs with [Ir(ppz)2(47-Ph2phen)][PF6] and[Ir(ppz)2(3478-Me4phen)][PF6] respectively increase to 5488 and 5236 cdm2
on blending with an ionic liquid [46] Similar observations are reported for LECswith [Ir(ppz)2(56-Me2phen)][PF6] [47]
Triazole-based cyclometallating ligands and the corresponding N^N ligands areconveniently prepared using lsquoClick chemistryrsquo methodology (Scheme 77) [48]An example of the effect of going from [ppy]ndash to the triazole-based [Phtl]ndash (seeScheme 76 for HPhtl) comes from Zysman-Colman and coworkers [49] Thecomplexes [Ir(Phtl)2(bpy)][PF6] and [Ir(Phtl)2(44prime-
tBu2bpy)][PF6] show PL emis-sion energies that are blue-shifted with respect to those of [Ir(ppy)2(bpy)][PF6] and[Ir(ppy)2(44prime-
tBu2bpy)][PF6] values are enhanced when the [Phtl]ndash ligand isintroduced The presence of the benzyl group (see Scheme 76) helps to stericallyprotect the iridium(III) centre and contributes to enhanced solution values of251 and 346 for [Ir(Phtl)2(bpy)][PF6] and [Ir(Phtl)2(44prime-
tBu2bpy)][PF6] com-pared to 929 and 271 for [Ir(ppy)2(bpy)][PF6] and [Ir(ppy)2(44prime-
tBu2bpy)][PF6]measured under the same room temperature conditions The synthesis of[Ir(C^N)2(N^N)]
+ complexes in which the C^N ligand contains a tetrazole unit (seeScheme 76 for the structure of Hprtz) proved to be challenging (see Sect 722)Isolated as the tetrafluoridoborate salts [Ir(prtz)2(bpy)]
+ [Ir(prtz)2(phen)]+ and
[Ir(prtz)2(44prime-tBu2bpy)]
+ exhibit intense PL bands with kemmax 540 nm and high
values of 55ndash70 Compared to complexes with triazole or pyrazole domainsthe use of a tetrazole-containing C^N ligand shifts the emission of the[Ir(C^N)2(N^N)]
+ complex even further to the blue Thin films (neat or in PMMAmatrix) of the complexes showed values of 27ndash70 with the best performingbeing [Ir(prtz)2(44prime-
tBu2bpy)][BF4] The EL spectrum of [Ir(prtz)2(44prime-tBu2bpy)]
+
is similar to that recorded under photoexcitation LEC devices with[Ir(prtz)2(44prime-
tBu2bpy)]+ in the emitting layer reach a maximum luminance of
310 cdm2 which is notably high for a blue-emitting device However the
Scheme 77 Synthesis of triazole-containing H(C^N) and N^N ligands using lsquoClick-chemistryrsquo(Huisgen cycloaddition) PMDTA = pentamethyldiethylenetriamine
178 CE Housecroft and EC Constable
luminance level drops to 20 cdm2 within 2 h and the instability of the device hasbeen attributed to unbalanced charge transport within the LEC [31]
The use of the phenylimidazoles Hdphim and Hbuphim (Scheme 76) ascyclometallating ligands in [Ir(C^N)2(N^N)]
+ luminophores was first reported in2016 [50] For comparisons with previously discussed compounds in this sectionwe focus on [Ir(dphim)2(bpy)][PF6] which has a solution PL kem
max of 591 nm and of 13 Compared to the corresponding energy levels in [Ir(ppy)2(bpy)][PF6]electrochemical data reveal that the HOMO and LUMO of [Ir(dphim)2(bpy)][PF6]are destabilized by 006 and 022 eV respectively Incorporated into a LEC withconfiguration ITOPEDOTPSSIr-iTMCAl [Ir(dphim)2(bpy)][PF6] attained avalue of Bmax = 503 cdm2 in 285 min (note that in this work ton = 28 min isdefined as the time to achieve B = 1 cdm2) The EL maximum of 578 nm isblue-shifted with respect to the solution PL kem
max (591 nm) The results suggest thatimidazole-based C^N ligands could be a useful addition to the colour-tuning paletteof ligands but device stability is not discussed [50]
734 N^N Ligands with Nitrogen-Rich Heterocycles
In this section we explore the effects of exchanging the bpy ligand in [Ir(ppy)2(bpy)]
+ by an N^N ligand containing one or two nitrogen-rich heterocycles(Scheme 78) The incorporation of additional nitrogen atoms into the N^N ligandis expected to shift the emission band of the [Ir(ppy)2(N^N)]
+ complex towards theblue because of the resulting increase in the HOMO-LUMO gap In 2008 Qiu andcoworkers demonstrated [51] that [Ir(ppy)2(pzpy)][PF6] exhibits solution PLkemmax = 475 nm which is considerably blue-shifted from the PL kem
max = 585 nm of[Ir(ppy)2(bpy)][PF6] LEC devices using [Ir(ppy)2(pzpy)]
+ show a blue-green EL(486 512 nm) and a maximum luminance of 52 cdm2 which increases to 94 cdm2
when [BMIM][PF6] is blended with the luminophore in the device (Ir-iTMCIL = 1035) The Effmax of the device (43 cdA) was reported at the time to be oneof the highest for a blue-green LEC [51] The t12 of 32 min was increased to218 min by replacing the cyclometallating [ppy]ndash ligands by thepyrazole-containing [ppz]ndash (see Scheme 76) but this was at the expense ofthe luminance (Bmax = 12 cdm2) [52] Choe and coworkers [53] have reportedthe performances of LECs containing [Ir(ppy)2(3-Mepzpy)]+ and[Ir(ppy)2(35-Me2pzpy)]
+ (see Scheme 78) with a configuration ITOPEDOTPSSIr-iTMCAl These LECs show Bmax of 351 and 549 cdm2 with yellow EL(kem
max = 537 and 543 nm respectively) The luminance can be enhanced by intro-ducing fluoro-substituents into the cyclometallating ligands (see Sect 75) [53] anda value of Bmax = 1246 cdm2 is reported on going to [Ir(ppy)2(Phpzpy)]
+ (seeScheme 78) with a bulky phenyl substituent in the pyrazole ring [54] However forthis series of LECs device stability was not discussed
Scheme 78 shows the structures of several N^N ligands involving imidazolemoieties The complex [Ir(ppy)2(EPimid)][PF6] (see Scheme 78) emits in the blue
7 Development of Cyclometallated Iridium(III) Complexes hellip 179
in solution (kemmax = 488 nm with shoulder at 511 nm) This is red-shifted in a thin
solid-film and the EL maximum is at 522 nm A LEC of configurationITOPEDOTPSSIr-iTMCAl in which the active layer contains[Ir(ppy)2(EPimid)]+ was slow to turn on but achieved a Bmax of 1191 cdm2 and aEffmax of 10 cdA The LEC was stable throughout the voltage scan to 12 V [55]The PL maximum of 554 nm for [Ir(ppy)2(pyim)][PF6] (in MeCN) is blue-shiftedwith respect to [Ir(ppy)2(bpy)][PF6] (585 nm) and a further blue-shift occurs ongoing to [Ir(ppy)2(bid)][PF6] (496 nm) Related complexes with benzimidazoleunits [Ir(ppy)2(Phpybi)]
+ and [Ir(ppy)2(qlbi)]+ exhibit kem
max at 588 and 627 nmrespectively (see Scheme 78 for structures of N^N ligands) For each complex theEL maximum is red-shifted by 10 nm LECs with configurations ITOPEDOTPSSIr-iTMC[BMIM][PF6]Al (Ir-iTMC[BMIM][PF6] = 1035 or 1040) areslow to turn-on (ton is defined in this case as the time to reach 1 cdm2) with turn-ontimes in the range 30ndash227 min times of between 29 and 86 h are needed to reachBmax The LEC with the fastest turn-on contains [Ir(ppy)2(bid)]
+ but its currentefficiency and EQE are low (04 cdA and 014) compared to the LECs containing[Ir(ppy)2(pyim)]+ (127 cdA and 42) and [Ir(ppy)2(Phpybi)]
+ (132 cdA and61) The poor performance of the LEC with [Ir(ppy)2(bid)]
+ is attributed to ahighly destabilized LUMO which impedes electron injection The stability of theseLECs was not discussed in detail [56] Increasing the length of the N-alkyl chain
Scheme 78 Structures and abbreviations of N^N ligands with nitrogen-rich heterocycles inSect 73
180 CE Housecroft and EC Constable
upon going from [Ir(ppy)2(Mepybi)]+ to [Ir(ppy)2(Etpybi)]+ to [Ir(ppy)2(Octpybi)]
+
(see Scheme 78) has little effect on the solution PL spectra with values ofkemmax = 572 579 and 576 nm respectively (yellow emissions) and values of 45
43 and 54 LECs were fabricated without ionic liquid (ITOPEDOTPSSIr-iTMCAl) and achieved maximum luminances of 2842 (N-methyl) 2933(N-ethyl) and 7309 cdm2 (N-octyl) indicating that the chain length has a significantimpact on device performance this is rationalized in terms of reduced inter-molecular interactions and self-quenching On going from [Ir(ppy)2(Mepybi)]+ to[Ir(ppy)2(Octpybi)]
+ the Effmax increases from 148 to 385 cdA Once againdevice stability was not discussed [57]
Luminophores [Ir(C^N)2(pytriaz-R)]+ combining fluoro-substituted cyclometal-
lating ligands with the triazole-containing ligand family pytriaz-R (Scheme 78)lead to blue-emitting LECs with fast turn-on times [58] The complex [Ir(ppy)2(pytetraz)][PF6] (see Scheme 78 for pytetraz) incorporates a tetrazole ringand exhibits a broad PL spectrum in solution with kem
max = 611 nm (orange-redemission) A blue-shift is observed on going to the EL spectrum (576 nm) A LECdevice (ITOPEDOTPSSIr-iTMCAl) with [Ir(ppy)2(pytetraz)]
+ in the active layerperforms poorly reaching a maximum luminance of only 66 cdm2 under a bias of9 V and it shows low stability under high voltages [59]
735 Designing N^N Ligands for Red-Emitting Iridium(III)Complexes
As of 2016 LECs exhibiting a deep-red emission were still poorly represented Toachieve a reduction in the HOMOndashLUMO gap with a consequential red-shift in theemission the HOMO of an Ir-iTMC must be destabilized andor the LUMO sta-bilized This can be achieved by introducing electron-donating groups onto the C^Nligands or electron-withdrawing groups onto the N^N ligand An alternativeapproach is to extend the aromatic system of the N^N domain for example byreplacing the 22prime-bipyridine ligand with a 22prime-biquinoline (biq Scheme 79) Thislatter approach was first reported by Thompson and coworkers [45] In MeCNsolution [Ir(ppz)2(biq)][PF6] exhibits a PL kem
max = 616 nm and this can be furtherred-shifted (to 627 nm) by introducing a tBu group into the phenyl ring of each[ppz]ndash ligand This latter complex was used in the active layer of a LEC of con-figuration ITOPEDOTPSSIr-iTMCAl and gave red EL (kem
max = 635 nm)The LEC reached a luminance of 1 cdm2 in 32 min after 70 min the peak EQEvalue of 74 was reached (Bmax = 7500 cdm2) This investigation was importantin demonstrating the potential for the use of N^N ligands with extended
7 Development of Cyclometallated Iridium(III) Complexes hellip 181
p-conjugation and has been followed by several related investigations The com-plex [Ir(ppy)2(biq)][PF6] has a solution PL maximum of 656 nm and has beencombined with blue-green emitters in a hostndashguest Ir-iTMCs configuration toproduce solid-state white light-emitting LECs [60] The ligand qlbi (Scheme 78) isrelated to biq and also possesses an extended aromatic system It has been incor-porated into [Ir(ppy)2(qlbi)]
+ which has a PL maximum at 627 nm with a shoulderat 646 nm In a LEC of configuration ITOPEDOTPSSIr-iTMC[BMIM][PF6](31)Al this Ir-iTMC shows a red EL at 650 nm but the turn-on time is long (72 hto reach 1 cdm2) and the maximum luminance is a moderate 70 cdm2 [61]Red-emitting [Ir(ppy)2(N^N)]
+ complexes containing 134-oxadiazole ligands(oxd Scheme 79) and giving high efficiencies in LECs have been reported [62]Thin films (spin-coated from CH2Cl2 solutions) containing [Ir(ppy)2(oxd)][PF6]with R = H or Me on the oxd ligand show PL kem
max values of 625 and 627 nmred-shifted from the solution maxima of 573 and 556 nm LECs with these com-plexes in the active layer and of configuration ITOPEDOTPSSIr-iTMC[BMIM][PF6] (41)Al were biased at 5 V and reached their maximum luminances of 217and 154 cdm2 in 72 and 35 min respectively The device lifetimes were 590 and490 min These results are notable in achieving an EQE value for a red-emittingLEC of 951 The fact that [Ir(ppy)2(4prime-tolyltpy)]
+ (4prime-tolyltpy = 4prime-(4-tolyl)-22prime6prime2primeprime-terpyridine see Scheme 710 for tpy) was known to exhibit orange-redPL was exploited by Constable et al in an investigation of [Ir(ppy)2(4prime-Xtpy)]
+
Ir-iTMCs [63 64] This is discussed in the next sectionThe strategy of introducing electron-withdrawing groups into the N^N ligand
has been demonstrated by incorporating [Ir(tolpy)2(44prime-(EtO2C)2bpy)][PF6] into aLEC (Htolpy = 2-tolylpyridine) In solution the complex emits at 687 nm whilstthe LEC exhibits EL at 630 nm The complex was also incorporated into a polymerand the polymer light-emitting electrochemical cell (PLEC) showed kem
max = 660nm Under a bias of 3 V the LEC turned on much faster than the PLEC (017 hversus 57 h to reach maximum luminance) consistent with lower anion mobility inthe latter Values of Bmax were 30ndash40 cdm2 for both devices but the PLECshowed an increased stability with respect to the small molecule-based LEC [65]There remains a need for long-lived deep-red emitters [66]
Scheme 79 Structures of some N^N ligands designed for use in red-emitting Ir-iTMCs
182 CE Housecroft and EC Constable
Scheme 710 Structures and abbreviations of some ligands including an aryl substituent adjacentto the metal-binding site
7 Development of Cyclometallated Iridium(III) Complexes hellip 183
74 Increasing Stability Through Intramolecularp-Stacking in [Ir(C^N)2(N^N)]
+ Luminophores
In Sect 732 we described the beneficial effects of functionalizing N^N ligands in[Ir(ppy)2(bpy)]
+ and [Ir(ppy)2(phen)]+ complexes with bulky substituents
However we omitted from this discussion the incorporation of aryl substituents inthe 66prime-positions of bpy or the 29-positions of phen In 2008 Bolink andcoworkers [67] demonstrated that LECs containing [Ir(ppy)2(ptbpy)]
+ (Fig 77)were particularly stable and long-lived due to the presence of an intramolecularface-to-face p-stacking interaction between the 6-phenyl substituent of the bpyligand and one of the cyclometallating rings The interaction was confirmed for theground state by X-ray crystallography (Fig 77) and DFT calculations showed thatthe stacking is also present in the excited state The structural figure in Fig 77shows the way in which the p-stacking interaction protects the iridium(III) centreand reduces the chance of attack by water molecules in the excited state byrestricting expansion along the NndashIrndashN vector resulting from the occupancy of anorbital with significant antibonding character between these atoms For a LECwith configuration ITOPEDOTPSSIr-iTMC[BMIM][PF6]Al (Ir-iTMC[BMIM][PF6] = 1026 41) under a bias of 3 V the time to reach the maximumluminance of 626 cdm2 was 6 h and the t12 was 25 h In contrast with a Ir-iTMC[BMIM][PF6] ratio of 1013 the turn-on time was 136 h and Bmax was 230 cdm2
and the t12 was estimated to be 680 h At the time of publication this repre-sented a record for LEC stability In the absence of IL a bias of 8 V was required toturn on a LEC with [Ir(ppy)2(ptbpy)]
+ in the active layer Parallel studies by Bolinket al [68] compared the performances of LECs with [Ir(ppy)2(bpy)]
+ or [Ir(ppy)2(Phbpy)]
+ (Fig 78) in the active layer and the results again confirmed theimportance of the intra-cation face-to-face p-stacking interaction A LEC of con-figuration ITOPEDOTPSS[Ir(ppy)2(Phbpy)][PF6]Al (with no ionic liquid)exhibited a turn-on time of several days consistent with low ion mobility This was
Fig 77 The structure of theN^N ligand 6-phenyl-4-(2-thienyl)-22prime-bipyridine(ptbpy) and thecrystallographicallydetermined structure of [Ir(ppy)2(ptbpy)]
+ in the [PF6]ndash
salt (CSD refcode NOLXES)[67]
184 CE Housecroft and EC Constable
overcome by adding ionic liquid to the active layer of the LEC and by also drivingthe device with pulsed high-voltage bias Under a bias of 7 V a maximum lumi-nance of 2200 cdm2 was reached within one minute thereafter theorange-emitting device was driven at 3 V and produced a value of Bmax =290 cdm2 with Effmax = 97 cdA and EQE = 4 The stability of the LEC wasexceptional with an extrapolated t12 of gt 3000 h compared to 30 h for a LECcontaining [Ir(ppy)2(bpy)]
+ Replacing Phbpy by Phphen (Scheme 710) has asimilar stabilizing influence [69] Interestingly however incorporating a secondp-stacking interaction by going from [Ir(ppy)2(Phbpy)]
+ to [Ir(ppy)2(66prime-Ph2bpy)]+
(Fig 78) did not lead to a further enhanced LEC performance although the LECwith [Ir(ppy)2(66prime-Ph2bpy)]
+ remains long-lived (t12 1300 h) [70]The strategy of using 6-phenyl-substituted bpy ligands in [Ir(C^N)2(N^N)]
+
luminophores has lead to some exceptional results in particular when this iscombined with the use of bulky peripheral substituents which inhibit closecationhellipcation approach in the active layer thus reducing exciton quenching Thisis nicely demonstrated by a comparison of the performances of LECs whichuse [Ir(ppy)2(Phbpy)]
+ [Ir(ppy)2(MeO-Phbpy)]+ [Ir(ppy)2(decylO-Phbpy)]+ or
[Ir(ppy)2(denO-Phbpy)]+ (see Fig 78 and Scheme 710) combined with ionic
liquid (Ir-iTMC[BMIM][PF6] = 4 1) in the active layer At a driving voltage of3 V the values of Bmax increase with steric demands of the peripheral group[Ir(ppy)2(decylO-Phbpy)]
+ (284 cdm2) gt [Ir(ppy)2(MeO-Phbpy)]+ (183 cdm2) gt[Ir(ppy)2(Phbpy)]
+ (109 cdm2) Device stability are in the range of t12 = 660ndash1290 h the longest being observed with [Ir(ppy)2(Phbpy)]
+ However even afterbeing held under a bias of 3 or 4 V for 48 h the LEC containing[Ir(ppy)2(denO-Phbpy)]
+ failed to turn on indicating that there is a limit to the sizeof the substituents that can be introduced denO-Phbpy contains a highly stericallydemanding dendritic domain (Scheme 710) [71] However one must be cautiousabout applying a simple sterics versus performance correlation Development ofligands for Ir-iTMCs goes hand-in-hand with changes and improvements in the
Fig 78 Structures of the ligands 6-phenyl-22prime-bipyridine (Phbpy) and 66prime-diphenyl-22prime-bipyridine (Ph2bpy) and the crystallographically determined structures of [Ir(ppy)2(Phbpy)]
+ inthe [PF6]
ndash salt [68] and [Ir(ppy)2(Ph2bpy)]+ in the [PF6]
ndash salt (CSD refcode COYMOT) [70]
7 Development of Cyclometallated Iridium(III) Complexes hellip 185
ways in which LECs are driven and the application of a pulsed current drivingmethod [72 73] has had an impact on turn-on times and also stabilizes the dopedregions in the device resulting in longer t12 Using such a pulsed driving mode aLEC containing [Ir(ppy)2(Ph2N-Phbpy)]
+ (Scheme 710) with the stericallydemanding 4-bis(35-(C6H4-4-NPh2)2phenyl group reached a luminance of75 cdm2 in 04 h and a final Bmax of 83 cdm2 and had a t12 of 250 h [74] It hasalso been concluded that the presence of bromo-substituents in the Ir-iTMC isdetrimental to LEC performance On going from [Ir(ppy)2(Ph2N-Phbpy)]
+ to[Ir(ppy)2(Br2-Phbpy)]
+ (Scheme 710) Bmax increases from 83 to 101 cdm2 butt12 decreases from 250 to 29 h [74] Exceptionally stable LECs have also beenachieved by replacing the 6-phenyl substituent in the bpy ligand by a 2-naphthylunit (Scheme 710) Again a pulsed driving mode was applied and a LECin configuration ITOPEDOTPSS[Ir(ppy)2(Naphbpy)][PF6][BMIM][PF6]Al(Ir-iTMC[BMIM][PF6] = 41) reached Bmax gt 300 cdm2 with no decay in lumi-nance over a 350 h period Significantly this bright and stable LEC was achievedeven though the solution was only 77 a fact that was explained in terms of thepresence of a low-lying triplet state associated with the 2-naphthyl unit which isclose to the MLCTLLCT lowest-energy emitting triplet [75]
Appending an aryl group adjacent to the metal-binding site is not restricted to theN^N ligand and aryl-substituted C^N ligands also lead to intra-cation p-stackinginteractions although with variable consequences Li et al have reported that LECsin configuration ITOPEDOTPSSIr-iTMC [BMIM][PF6]Al (Ir-iTMC[BMIM][PF6] = 41) driven under a bias of 3 V and with [Ir(26-Ph2py)2(phen)]
+ or[Ir(246-Ph3py)2(phen)]
+ in the active layer do not turn on even after 24 hincreasing the bias to 8 V did not lead to light emission The double p-stack in the[Ir(246-Ph3py)2(phen)]
+ cation is shown on the left in Fig 79 and leads to the phenligand being sandwiched between two phenyl rings in turn this causes the iridium(III) coordination environment to be noticeably distorted DFTTD-DFT calculationsreveal that in both [Ir(26-Ph2py)2(phen)]
+ and [Ir(246-Ph3py)2(phen)]+ the
p-stacking contacts do not prevent the iridium(III) coordination sphere from openingup in the 3MC state [76] This is in contrast to the beneficial effects seen when thearyl substituent is attached to the 6-position of bpy and the p-stacking interaction isbetween the pendant aryl ring and a cyclometallated ring (see Figs 77 and 78)With this in mind the ligand 2-(35-diphenyl)phenylpyridine (H(Ph2ppy)Scheme 710) was designed to give two separate p-stacking interactions in thecomplex [Ir(Ph2ppy)2(bpy)]
+ the two cyclometallating ligands embrace one anotheras shown in the centre of Fig 79 A LEC with a blend of [Ir(Ph2ppy)2(bpy)][PF6]and [BMIM][PF6] (4 1) in the active layer and operated under a pulsed currentdriving mode reached a maximum luminance of 425 cdm2 in 121 h and the devicewas very stable (t12 = 360 h) Paradoxically however a much brighter(Bmax = 1024 cdm2) and more stable (t12 = 2800 h) LEC was fabricated using[Ir(Phppy)2(bpy)][PF6] The H(Phppy) ligand lacks the pendant phenyl ring ofH(Ph2ppy) and as a result there are no intra-cation p-stacking contacts in[Ir(Phppy)2(bpy)]
+ (Fig 79 right) Significantly a combination of a pendant6-phenyl ring in the bpy domain with either H(Phppy) or H(Ph2ppy) as the
186 CE Housecroft and EC Constable
cyclometallating ligand leads to highly efficient LECs the performance parameters ofdevices with [Ir(Phppy)2(Phbpy)][PF6] or [Ir(Ph2ppy)2(Phbpy)][PF6] (and IL) in theemissive layer are summarized in Table 73 along with those for LECs with[Ir(Phppy)2(bpy)][PF6] and [Ir(Ph2ppy)2(bpy)][PF6] [35] The exceptional stability(Table 73) of the device with [Ir(Ph2ppy)2(Phbpy)]
+ in the active layer is assumed tobe associated with the face-to-face p-stacking interactions shown in Fig 710 whichminimize attack at the metal centre by nucleophiles Taking all the data discussed inthis section into account it appears that the p-contact between the phenyl ring of thePhbpy ligand and one of the cyclometallating rings plays a key role
The design of [Ir(C^N)2(N^N)]+ luminophores involving p-stacking between a
pendant phenyl ring of the N^N ligand and one cyclometallating ring is indeed asuccessful strategy for achieving stable LECs A comparison of the performances of
Fig 79 Left the structure of the [Ir(246-Ph3py)2(phen)]+ cation from the [PF6]
ndash salt (CSDrefcode YODFII) [76] Middle the structure of [Ir(Ph2ppy)2(bpy)]
+ cation from the [PF6]ndash salt
(CSD refcode BOYPAI) with the two C^N ligands shown in red and pale blue respectively [35]Right the structure of the [Ir(Phppy)2(bpy)]
+ cation from the [PF6]ndash salt (CSD refcode BOYPEM)
[35] In each figure the N^N ligand (phen or bpy) is on the right-hand side and p-stackinginteractions are shown in space-filling representations
Table 73 Performances of LECs (ITOPEDOTPSSIr-iTMC [BMIM][PF6] (41)Al) containingmultiple phenyl substituents on the C^N and N^N domains in [Ir(C^N)2(N^N)]
+ Phbpy H(Phppy)and H(Ph2ppy) are defined in Fig 78 and Scheme 710 Each complex was a [PF6]
ndash salt anddevices were driven by applying a pulsed current density of 300 Am2 Data from Ref [35]
Luminophore ton (h)a Bmax (cdm
2) Effmax (cdA) t12 (h)b
[Ir(Phppy)2(bpy)]+ 014 1024 35 2800
[Ir(Ph2ppy)2(bpy)]+ 121 425 14 360
[Ir(Phppy)2(Phbpy)]+ 042 676 22 1204
[Ir(Ph2ppy)2(Phbpy)]+ 005 261 07 gt2800
a ton = time to reach Bmaxb t12 = time for luminance to decay from Bmax to Bmax2
Note The terms ton t12 Bmax Effmax refer to turn-on time lifetime maximum luminancemaximum efficiency respectively The definition of these terms are provided in Chap 1
7 Development of Cyclometallated Iridium(III) Complexes hellip 187
LECs containing [Ir(ppy)2(Phbpy)]+ (Fig 78) or [Ir(dmppz)2(Phbpy)]
+ (Fig 711left) in configuration ITOPEDOTPSSIr-iTMC[BMIM][PF6]Al (Ir-iTMC[BMIM][PF6] = 41) and under a 3 V bias confirms that the device stabilityobserved for [Ir(ppy)2(Phbpy)]
+ (discussed earlier in this section) is replicated in[Ir(dmppz)2(Phbpy)]
+ The t12 is further increased (from 1288 to 2000 h on goingfrom [Ir(ppy)2(Phbpy)]
+ to [Ir(dmppz)2(Phbpy)]+) by the effects of the methyl
substituents in the pyrazole ring which hinder the approach of nucleophiles [77] Apendant phenyl substituent adjacent to the metal-binding site is also effective whenthe N^N ligand is N-Phpypz (Fig 711) The [Ir(ppy)2(N-Phpzpy)]
+ cation exhibitsthe expected stacking of the pendant phenyl group and one cyclometallating ring(Fig 711 right) The effect of introducing the phenyl substituent was assessed bycomparing LECs containing [Ir(ppy)2(N-Phpzpy)][PF6] and [Ir(ppy)2(pzpy)][PF6](see Scheme 78 for pzpy) In a configuration ITOPEDOTPSSIr-iTMCBMIMPF6Al (Ir-iTMC[BMIM][PF6] = 21) and under a 3 V bias the LEC with[Ir(ppy)2(pzpy)][PF6] reached a maximum luminance of 14 cdm2 in 38 min andhad a t12 of 300 min When the pendant phenyl group is introduced ton Bmax
and t12 of the blue-green LEC dramatically increase (325 min to reach 37 cdm2and t12 = 950 min) [78] However He et al note that at the time of theirachievement (2011) the brightness and t12 of this blue-green LEC remain signif-icantly lower than those of the most stable orange-red LECs then known Whilstfluoro-substituents in the cyclometallating ligands are used to realize blue emitters(see Sect 75) Chen et al have sought to enhance intra-cation p-stacking by usingH(dfppz) which contains an electron-poor difluorophenyl cyclometallatingring (Fig 712) Each of [Ir(dfppz)2(sp)][PF6] [Ir(dfppz)2(Phsp)][PF6] or[Ir(dfppz)2(Ph2sp)][PF6] was blended with ionic liquid ([BMIM][PF6]) in the activelayer of a LEC The t12 undergoes a 4-fold increase on going from [Ir(dfppz)2(sp)]
+
to [Ir(dfppz)2(Phsp)]+ increasing from 141 min to 585 min but introducing two
phenyl rings (and therefore two p-stacking interactions) does not have a beneficialeffect (t12 = 102 min) [79] This contrasts with the fact that LECs with both
Fig 710 The three face-to-face p-stacking interactions in [Ir(Ph2ppy)2(Phbpy)]+ There are two
pyridylhellippendant phenyl interactions (shown in pale blue and orange) and one face-to-facep-contact between the phenyl ring of the Phbpy ligand and one cyclometallating ring (shown inred) Data from CSD refcode BOYPIQ [35]
188 CE Housecroft and EC Constable
[Ir(ppy)2(Phbpy)]+ (one p-stacking contact) and [Ir(ppy)2(66prime-Ph2bpy)]
+ (twop-stacking interactions) are long-lived even though there is no advantage to therebeing an additional p-stacking interaction (see above) [70]
Structurally 22prime6prime2primeprime-terpyridine (tpy Scheme 710) is closely related to6-phenyl-22prime-bipyridine (Phbpy) Although tpy is most often encountered as a bis(chelating) NNprimeNprimeprime-donor ligand [80] it can only act as a bidentate ligand in an[Ir(C^N)2(tpy)]
+ complex because of the restriction to 6-coordination at the metalcentre One pyridyl ring in the bidentate tpy in [Ir(ppy)2(tpy)]
+ therefore mimics thependant phenyl ring in [Ir(ppy)2(Phbpy)]
+ As expected complexes of the type[Ir(ppy)2(Xtpy)]
+ (X represents a peripheral functional group) show face-to-facep-stacking of the pendant pyridyl ring with one of the cyclometallating rings Althoughthis appeared promising for the design of luminophores for LECs and despite the factthat the emissions were in the red-orange region [Ir(ppy)2(Xtpy)][PF6] complexesshowed values of in solution or thin films of lt25 and poor performances in LECs[63 64]
75 Effects on LEC Stability of IntroducingFluoro-Substituents into Cyclometallating Ligands
As discussed in Sect 71 colour tuning of the emission wavelength of a[Ir(C^N)2(N^N)]
+ luminophore is readily achieved by making use of the parti-tioning of IrC^N vs N^N character in the HOMO and LUMO respectively
Fig 711 Left the structure of the cyclometallating ligand Hdmppz and of the[Ir(dmppz)2(Phbpy)]
+ cation (CSD refcode YUWWOD) [77] Right the N^N ligand N-Phpzpyand the [Ir(ppy)2(N-Phpzpy)]
+ cation (CSD refcode UTADAV) [78]
7 Development of Cyclometallated Iridium(III) Complexes hellip 189
Thus lowering the energy of the HOMO by introducing electron-withdrawingfluoro groups into the cyclometallating ligands is regularly used to push theemission towards the blue This topic is discussed in detail in Chap 8 Here wefocus on how LEC stability is affected when fluoro-substituents are introduced intothe cyclometallating ligands The fluoro-substituted ligands discussed in this sectionare shown in Scheme 711
An important contribution to quantifying the effects that fluoro-substitution canhave on LEC lifetimes comes from the Bolink and Frey groups [81] The number ofF atoms and the substitution pattern in the cyclometallating rings was varied in theseries [Ir(5-Fppy)2(44prime-
tBu2bpy)]+ and [Ir(3-Fppy)2(44prime-
tBu2bpy)]+ [Ir(dfppy)
(ppy)(44prime-tBu2bpy)]+ and [Ir(Me-dfppy)2(44prime-
tBu2bpy)]+ (see Schemes 75 and
711) Each of the first three complexes contains two F atoms whilst the lastcomplex contains four Despite this their photophysical and electrochemicalproperties were essentially the same (solution PL kem
max = 552 to 555 nmEox = + 100 to +103 V versus FcFc+ compare Table 71) thus allowing a validcomparison of LEC device characteristics LECs in configuration ITOPEDOTPSSIr-iTMC[BMIM][PF6]Al (Ir-iTMC[BMIM][PF6] = 41) were biased with apulsed current and featured short turn-on times and similar values of Bmax
(1028ndash1095 cdm2) However whereas LECs with the luminophores containingtwo F atoms had t12 values in the range 483ndash598 h that with[Ir(Me-dfppy)2(44prime-
tBu2bpy)]+ (four F atoms) exhibited a shorter t12 of 132 h
Fig 712 Schematic structures of luminophores containing fluoro-substituted C^N ligands anddiazaspirobifluorene ancillary ligands and the p-stacking interaction in [Ir(dfppz)2(Phsp]
+ (CSDrefcode GEMXAZ) [79]
Scheme 711 Structures of fluoro-substituted cyclometallating ligands
190 CE Housecroft and EC Constable
A number of investigations have compared performances of fluoro-containingand non-fluoro-containing LECs although in some cases direct comparisons aredifficult because of other structural variations In Sect 734 we described theeffects of increasing the N-alkyl chain upon going from [Ir(ppy)2(Mepybi)]+ to[Ir(ppy)2(Etpybi)]
+ to [Ir(ppy)2(Octpybi)]+ (Scheme 712) Bmax and the Effmax
increased [57] Introducing fluoro-substituents (Scheme 712) shifted the ELemission maximum from the yellow (575 574 and 566 nm for the three complexeswith [ppy]ndash ligands) to the green (527 529 and 527 nm for luminophores with[dfppy]ndash ligands) but had a detrimental effect on luminance (Bmax decreased from2842 to 1520 cdm2 for R = Me from 2933 to1642 cdm2 for R = Et and from7309 to 3112 cdm2 for R = noctyl) The device stability was not discussed in detail[57] A similar reduction in luminance was observed for LECscontaining [Ir(ppy)2(Phpzpy)]
+ (see Scheme 78 Bmax = 1246 cdm2) versus[Ir(dfppy)2(Phpzpy)]
+ (Bmax = 674 cdm2) [54] and for [Ir(ppy)2(EPimid)]+ (seeScheme 78 Bmax of 1191 cdm2) versus [Ir(dfppy)2(EPimid)]+ (Bmax of741 cdm2) [55] the active layers in these devices contained no ionic liquid and thet12 were not detailed In contrast LECs with [Ir(dfppy)2(3-Mepzpy)]+ or[Ir(dfppy)2(35-Me2pzpy)]
+ (see Schemes 78 and 711) in the active layer arereported to have enhanced luminance and Effmax compared to analogous LECs with[Ir(ppy)2(3-Mepzpy)]+ or [Ir(ppy)2(35-Me2pzpy)]
+ The superior film-formingproperties of [Ir(dfppy)2(35-Me2pzpy)][PF6] spin-coated from an acetonitrilesolution contributed to the performance of this LEC [53] hence the direct effectsof introducing fluoro-substituents are difficult to assess
Replacing the cyclometallating [ppy]ndash in an [Ir(ppy)2(N^N)]+ complex by either
[dfppy]ndash or [ppz]ndash leads to blue-shifted emission maxima (see Sect 733 for the useof Hppz) and Scheme 713 shows three series of [Ir(dfppy)2(N^N)]
+ and[Ir(ppz)2(N^N)]
+ complexes that have been the subject of comparative investiga-tions [82ndash84] In each case LECs had a configuration ITOPEDOTPSSIr-iTMCAl (no ionic liquid) For the pair of luminophores[Ir(dfppy)2(5-Mephen)]+ and [Ir(ppz)2(5-Mephen)]+ the latter had a lower turn-onvoltage (9 V) to reach Bmax of 1549 cdm2 compared to a bias of 10 V required for[Ir(dfppy)2(5-Mephen)]+ to achieve Bmax = 2430 cdm2 The current density for the
Scheme 712 Series of complexes investigated [57] for the influence of N-alkyl chain length inthe N^N ligand and influence of fluoro-substitution in the C^N ligand
7 Development of Cyclometallated Iridium(III) Complexes hellip 191
LEC with [Ir(dfppy)2(5-Mephen)]+ was 8483 mAcm2 compared to a value of4571 mAcm2 for the LEC with [Ir(ppz)2(5-Mephen)]+ The latter LEC has a lowstability under high bias Similarly a LEC with [Ir(dfppy)2(29-
nBu2phen)]+ in the
active layer showed a higher luminance (Bmax = 947 cdm2) than a device con-taining [Ir(ppz)2(29-
nBu2phen)]+ (Bmax = 773 cdm2) and for both the bias
required to turn on the devices was high (10 V) Long-term stabilities were notdiscussed The trend was similar for a pair of LECs with [Ir(dfppy)2(N-Hpyim)]+
and [Ir(ppz)2(N-Hpyim)]+ but luminance levels were lowThe effects of introducing the bulky and strongly electron-withdrawing SF5 group
into the cyclometallating ligandsHppy andHppz have also been investigatedAlthoughthe [Ir(C^N)2(44prime-
tBu2bpy)]+ complexes in which HC^N is one of the ligands shown
Scheme 714 exhibit PL in solution and in thin films they do not show EL in LECswhich were driven using a pulsed current method It was proposed that the chemicalinstability of the SF5 group is responsible for the lack of device performance [85]
76 Fluorine-Free Blue-Shifted Emitters
The use of a sulfonyl group such as MeSO2 to replace the electron-withdrawingfluorine in the cyclometallating ligand has been shown theoretically to be viable forshifting the emission of the Ir-iTMC towards the blue [86] Scheme 715 showssome of the methylsulfonyl-functionalized ligands investigated to date The use ofHmsppz combines the effects of the electron-withdrawing MeSO2 group with theelectron-rich pyrazole ring both of which widen the HOMOndashLUMO gap Pulseddriving of LECs containing [Ir(msppz)2(N^N)]
+ (N^N = six functionalized bpy
Scheme 713 Three series of [Ir(dfppy)2(N^N)]+ and [Ir(ppz)2(N^N)]
+ complexes the N^Nligands are defined on the right
Scheme 714 Cyclometallating ligands containing electron-withdrawing SF5 substituents
192 CE Housecroft and EC Constable
ligands) combined with [BMIM][PF6] in the active layer led to fast turn-on timesof the order of seconds for these green-emitting devices High luminances wereobserved (up to 1127 cdm2) but the t12 values were very short (02 to 54 min)This result suggested that the presence of fluorine atoms in the luminophore is onlyone contributing factor for green or blue-green-emitting LECs being short-lived[86] An extension of this work led to an evaluation of [Ir(5-msppy)2(bpy)]
+ [Ir(4-msppy)2(bpy)]
+ and [Ir(3-msppy)2(bpy)]+-based emitters (see Scheme 78 for
pypz) and of [Ir(5-msppy)2(pypz)]+ and [Ir(4-msppy)2(pypz)]
+ in LECs of config-uration ITOPEDOTPSSIr-iTMC[BMIM][PF6]Al (Ir-iTMC[BMIM][PF6] = 41) which were operated using a pulsed current driving mode [28 87]The position of substitution of the sulfonyl group in the cyclometallated ring in the[Ir(msppy)2(bpy)]
+ series has a significant effect on the maximum luminance(4-msppy gt 3-msppy 5-msppy) Despite the brightest luminance being940 cdm2 all the LECs are short-lived ( 15 h) LECs with [Ir(5-msppy)2(pypz)]
+ and [Ir(4-msppy)2(pypz)]+ in the active layer also show low
stability with values of t12 of between 24 and 62 min in this series maximumluminances are also low ( 141 cdm2) Although the introduction of methylsul-fonyl groups does not appear to aid device stability positional isomerism has beenshown to be an effective means of colour tuning the emission [28 87]
An alternative strategy for fluorine-free blue-green emitters comes fromZysman-Coleman andBolink [88]A combinationof a cyclometallated pyridyl (ratherthan phenyl) ring with electron-donating functionalities (in this case methoxy) hasbeen shown to be electronically similar to using a fluoro-substituted cyclometallated[dfppy]ndash ligand Turn-on times for devices driven under pulsed current conditionswere very fast (lt1 s) and whilst the best performing LEC achievedBmax = 1054 cdm2 the LECs suffered from poor stability with values of t12 ofbetween 0007 and 246 h The search for alternatives tofluoro-substituents continues
77 Replacing the N^N Ligand in [Ir(C^N)2(N^N)]+
Emitters by N-Heterocyclic Carbenes
The use of N-heterocyclic carbenes (NHCs) as strong-field ligands to replace theN^N ligand in [Ir(C^N)2(N^N)]
+ complexes in LECS was first reported by De Colaand coworkers [89] The aim of shifting the emission towards the deep-blue
Scheme 715 Cyclometallating ligands containing electron-withdrawing methylsulfonylsubstituents
7 Development of Cyclometallated Iridium(III) Complexes hellip 193
was achieved with the luminophores [Ir(dfppy)2(Me-mdiim)]+ and [Ir(dfppy)2(Bu-mdiim)]+ (see Schemes 711 and 716) which were used as either the[PF6]
ndash or [BF4]ndash salts in LECs of configuration ITOPEDOTPSSIr-iTMCTBATf
(11)Al The devices featured the new ionic conductor tetrabutylammonium triflate(TBATf) and were driven using either a voltage sweep or constant voltage modeA shorter N-alkyl chain in the NHC leads to brighter emission with the highestvalue of Bmax being observed for the LEC with [Ir(dfppy)2(Me-mdiim)]+
(1598 cdm2 for the [BF4]ndash salt and 698 cdm2 for the [PF6]
ndash salt) the times toreach these luminances were 106 and 45 min respectively The stability of theLECs was relatively low with t12 between 80 and 538 min A series of [Ir(C^N)2(pybzim)]+ complexes in which the cyclometallating ligands include HppyHdfppy and H(dfpypy) (see Schemes 711 716 and 717) showed poor photo-physical properties and DFTTD-DFT calculations supporting this experimentalstudy showed that the PL originates from the C^N ligand rather than the NHC Thiscontrasts with the more commonly employed [Ir(C^N)2(N^N)]
+ emitters in whichemission originates from the N^N ligand The relatively low s for the [Ir(C^N)2(pybzim)]+ complexes was attributed to non-radiative deactivation of theexcited states by thermally accessible iridium-centred states [90] When used in the
Scheme 716 Structures of N-heterocyclic carbenes
Scheme 717 Cyclometallating ligands used in combination with NHCs in Ir-iTMCs
194 CE Housecroft and EC Constable
active layers of ITOPEDOTPSSIr-iTMC[BMIM][PF6] (31)Al LECs under a6 V bias [Ir(ppy)2(pybzim)][PF6] and [Ir(dfppy)2(pybzim)][PF6] gave only lowluminances (Bmax 20 cdm2) and the EL maxima (544 and 512 nm) werered-shifted by with respect to the PL (kem
max = 471 and 450 nm respectively) Incontrast a LEC containing [Ir(4-MeO2C-2-Phpy)2(pybzim)]+ (see Scheme 717)produced a bright orange emission (Bmax = 1070 cdm2 EL kem
max = 584 nm) withan Effmax of 47 cdA The stabilities of the devices were not discussed [90]The NHC Me-mdiim has also been combined with the cyclometallating ligand[dfpypytBu]ndash (Scheme 717) with the result that [Ir(dfpypytBu)2(Me-mdiim)][PF6]gives a deep-blue emission at 440 nm in MeCN solution ( = 13 s = 116 ls)[Ir(dfpypytBu)2(Me-mdiim)][PF6] was used in the active layer of a LEC of con-figuration ITOPEDOTPSSIr-iTMC[BMIM][PF6] (31)Al The EL spectrum wastime dependent when the device was biased at 6 V the initial EL spectrum(kem
max 500 nm) slowly develops a tail into the red and after 10 min kemmax =
532 nm with the profile of the EL spectrum undergoing significant changeA change from a driving mode of 6 V fixed bias to 25 mAcm2 constant current hasa beneficial effect on the LEC luminance (15 cdm2 increased to 113 cdm2) but ledto slower turn-on (215 s versus 17 s) However the LEC has a very short lifetime(t12 of 179 or 60 s depending on the operating mode) DFTTD-DFT calculationsagain (see above) indicate that the PL originates from the C^N ligand rather than theNHC [91] In an attempt to improve stability an N-phenyl substituent has beenintroduced into the NHC ligand pyphmi with the aim of replicating the advanta-geous intra-cation p-stacking discussed in Sect 74 Crystallographic data for[Ir(ppy)2(pyphmi)][PF6] confirmed the intended face-to-face interaction between thependant phenyl ring and one of the cyclometallating rings in the solid-state(Fig 713) However LECs using [Ir(ppy)2(pyphmi)][PF6] in the active layer didnot show enhanced stability with respect to these containing [Ir(ppy)2(pymemi)][PF6] on going from pyphmi to pymemi the pendant phenyl ring was replaced by a
Fig 713 Structure of the [Ir(ppy)2(pyphmi)]+ cation in [Ir(ppy)2(pyphmi)][PF6] (CSDrefcode MEGLUH) [92]
7 Development of Cyclometallated Iridium(III) Complexes hellip 195
methyl group (Scheme 716) [92] The results of theoretical studies indicate thatpopulation of the metal-centred 3MC excited states leads to rupture of the IrndashN bondto the pyridine ring of the HNC ligand [93] The results suggest that the intra-cationp-stacking which is so successful in [Ir(C^N)2(N^N)]
+ species will not prove such auseful design strategy in [Ir(C^N)2(HNC^N)]
+-type complexes
78 Effects of Incorporating Peripheral Charged Domainsin Ir-iTMCs and the Design of Anionic Ir-iTMCs
A cationic Ir-iTMC complex is typically paired with a large anion such as [PF6]ndash
which impedes ionic mobility and results in a slow turn-on time ton (Weemphasize here again that turn-on times are in some works defined as the time toreach maximum luminance ton or tmax and in others as the time to reach a specifedluminance ton) A number of strategies have been tried in attempts to decreasevalues of ton including the incorporation into Ir-iTMCs of ligands with positivelycharged peripheral groups (Scheme 718) An early study from Bolink et al usedthe positively charged cyclometallating ligands [H(ppyPBu3)]
+ (Scheme 718) so asto achieve a charged homoleptic [Ir(C^N)3]
3+ complex suitable for use in LECsthis contrasted with the more usual neutral homoleptic [Ir(C^N)3] complexes usedin organic light-emitting devices (OLEDs) In acetonitrile solution The PL kem
max isshifted from 515 to 480 nm on going from [Ir(ppy)3] to [Ir(ppyPBu3)3]
3+ while the suffers a small decrease (40ndash34) The efficiency and stability of a LEC
Scheme 718 H(C^N) and N^N ligands with positively charged peripheral groups and a referenceligand dC6-daf (see text)
196 CE Housecroft and EC Constable
(ITOIr-iTMCPMMAAu) was low but the results confirmed the principle of usingpositively charged ligands in the design of cyclometallated iridium(III) complexesfor LECs [94]
The family of 5-(x-triethylammonioalkyl-22prime-bipyridine bromides contain thecations [bpy(CH2)xNEt3]
+ and the cation with the shortest chain is shown inScheme 718 These ligands have been used to prepare dicationic Ir-iTMCs [Ir(ppy)2bpy(CH2)xNEt3]
2+ with the increased charge aimed at increasing ionicconductivity and thereby reducing LEC turn-on times LECs with configurationITOIr-iTMCAu were operated under a bias of 3 V Compared to a device with [Ir(ppy)(44prime-tBu2bpy)][PF6] in the active layer those with [Ir(ppy)2bpy(CH2)xNEt3][PF6]2 reached maximum luminance far faster (33ndash38 min for thealkyl-tailed luminophores compared to 940 min for [Ir(ppy)(44prime-tBu2bpy)]
+)However with a triethylammonioheptyl chain device-quality thin films could notbe made indicating a limitation imposed by the chain length The improved turn-ontime was unfortunately not matched by a corresponding enhancement of the sta-bility with a value of t12 = 988 min for the LEC with [Ir(ppy)(44prime-tBu2bpy)]
+
greatly exceeding the 54ndash84 min observed for the [Ir(ppy)2bpy(CH2)xNEt3]2+
series [95] A significant decrease in the turn-on time was also reported when [Ir(ppy)2(dC6MIM-daf)]3+ (see Scheme 718) was used in the active layer of a LECDevices in configuration ITOIr-iTMCAg with either [Ir(ppy)2(dC6MIM-daf)][PF6]3 or [Ir(ppy)2(dC6-daf)][PF6] in the active layer were operated under constantbias of 27 or 28 V A red-shift in the EL maximum from 566 to 577 nmaccompanied the introduction of the positively charged N-methylimidazoliumgroups At a bias of 28 V values of Bmax of 105 versus 79 cdm2 (without and withthe imidazolium units respectively) were reached in 500 versus 200 min The EQEof 5 was little altered on going from [Ir(ppy)2(dC6-daf)][PF6] to [Ir(ppy)2(dC6MIM-daf)][PF6]3 the long-term stabilities of the LECs was not dis-cussed [96]
In 2009 Hong and coworkers reported the salt shown at the top of Scheme 719In their design strategy an anionic ligand replaced the more typical N^N ancillaryligand and the introduction of an alkylsulfonate tail provided the necessary overallcharge for the complex A sodium counter-ion was chosen to enhance ionicmobility in the LEC The performances of LECs containing [Ir(pquin)2(bpy)][PF6]or Na[Ir(pquin)2(picSO3)] in the active layer were compared using a LEC con-figuration of ITOIr-iTMCPEOAu (PEO = polyethylene oxide) Values of the ELkemmax of 600 and 640 nm for [Ir(pquin)2(bpy)][PF6] and Na[Ir(pquin)2(picSO3)]
respectively are red-shifted with respect to the solution PL kemmax of 550 and
570 nm The LEC containing [Ir(pquin)2(bpy)][PF6] required a higher bias (54 V)to turn on than that with Na[Ir(pquin)2(picSO3)] in the active layer (36 V) TheLECs exhibited maximum luminances of 1213 and 990 cdm2 respectively and thedevice with Na[Ir(pquin)2(picSO3)] continuously emitted light as the voltageincreased The most significant finding in this study was that the turn-on time(defined in this case as the time to reach a luminance of 1 cdm2) diminished from30 to 05 min on replacing [Ir(pquin)2(bpy)][PF6] by Na[Ir(pquin)2(picSO3)] as theelectroluminophore Despite this promising result the Na[Ir(pquin)2(picSO3)]-
7 Development of Cyclometallated Iridium(III) Complexes hellip 197
containing LEC displayed a low efficiency and stability [97] In contrast to theanionic charge residing on a pendant chain Thompson and coworkers [98] designedthe two anionic Ir-iTMCs shown at the bottom of Scheme 719 A problem facinganionic luminophores is their high LUMO energy the LUMO levels of [Ir(pquin)2(CN)2]
ndash and [Ir(Meppy)2(CN)2]ndash are well above the work function of Al
leading to a high barrier for electron injection Poor solubilities of the Na[Ir(pquin)2(CN)2] and Na[Ir(Meppy)2(CN)2] in MeCN using during spin-coating thecomplexes onto the ITO substrate was overcome by using the crown ether18-crown-6 as an additive LECs with these anionic luminophores were fabricated inan ITOIr-iTMCAl configuration were short-lived and a contributing factor wasthought to be the presence of monodentate cyanido ligands which give no protectionto the iridium(III) centre in the excited state (contrast the use of the p-stackinginteractions described in Sect 74) Related complexes in which the cyanido ligandswere replaced by tetrazolates were reported in 2016 [99] and the lability of thetetrazolates was overcome by incorporating them into a bidentate ligand 12-H2BTB(Scheme 719c) [100] This structural design strategy appears to be a promising stepforward although the performance of these Ir-iTMCs in LECs has yet to be proven
Scheme 719 Structures of the sodium salts of the Ir-iTMCs (top) [Ir(pquin)2(picSO3)]ndash [Ir
(pquin)2(CN)2]ndash and [Ir(Meppy)2(CN)2]
ndash (bottom left) as well as 12-H2BTB (bottom right)
198 CE Housecroft and EC Constable
79 Conclusions
In this chapter we have provided an overview of the development of cyclometal-lated iridium(III) complexes for application in LECs The aim is to highlight theligand-design features that have helped to tune the emission colour and to improvedevice stability and operating efficiency The time for a LEC to achieve maximumluminance is a major hurdle that must be overcome if these devices are to reach thecommercial market and we have illustrated a number of approaches that have beenused to reduce ton One of the difficulties of reviewing the LEC literature is incomparing device data from different sources In particular the LEC configurationsare variable (cathode material use of hole injection material blending the Ir-iTMCwith ionic liquid) as are the operating conditions making direct comparisons ofLEC performance data difficult Nonetheless this overview allows the reader togain valuable insight into the evolution of state-of-the-art Ir-iTMC LECs and thedesign strategies that lie behind this progress
Acknowledgements We are grateful to the Swiss National Science Foundation The EuropeanUnion the Swiss Nanoscience Institute and the University of Basel for the generous support whichhas allowed us to make our own contributions to this exciting and topical area of materialschemistry and device engineering where appropriate individual grants are cited in references Weare indebted to the efforts of members of our research group whose contributions to the field ofIr-iTMC-based LECs are acknowledged in the references below Structural figures have beendrawn using data retrieved from the Cambridge Structural Database (CSD) using Conquest v 18[101] and Mercury v 36 [101 102] relevant structure reference codes (refcodes) are given infigure captions
References
1 Q Pei G Yu C Zhang Y Yang AJ Heeger Science 269 1086 (1995)2 KM Maness RH Terrill TJ Meyer RW Murray RM Wightman J Am Chem Soc
118 10609 (1996)3 RD Costa E Ortiacute HJ Bolink Pure Appl Chem 83 2115 (2011)4 JD Slinker J Rivnay JS Moskowitz JB Parker S Bernhard HD Abruntildea GG
Malliaras J Mater Chem 17 2976 (2007)5 MS Lowry S Bernhard Chem Eur J 12 7970 (2006)6 E Holder BMW Langeveld US Schubert Adv Mater 17 1109 (2005)7 C Ulbricht B Beyer CA Friebe A Winter US Schubert Adv Mater 21 4418 (2009)8 L Hu G Xu Chem Soc Rev 39 3275 (2010)9 T Hu L He L Duan Y Qiu J Mater Chem 22 4206 (2012)
10 RD Costa E Ortiacute HJ Bolink F Monti G Accorsi N Armaroli Angew Chem Int Ed51 8178 (2012)
11 KPS Zanoni RL Coppo RC Amaral NY Murakami Iha Dalton Trans 44 14559(2015)
12 RD Costa Mol Work 339 (2012)13 L Zhou S Liu Q Zhao Q Ling W Huang Prog Chem 23 1871 (2011)14 SB Meier D Tordera A Pertegaacutes C Roldaacuten-Carmona E Ortiacute HJ Bolink Mater Today
17 217 (2014)
7 Development of Cyclometallated Iridium(III) Complexes hellip 199
15 ST Parker JD Slinker MS Lowry MP Cox S Bernhard GG Malliaras ChemMater 17 3187 (2005)
16 JD Slinker CY Koh GG Malliaras MS Lowry S Bernhard Appl Phys Lett 86article number 173506 (2005)
17 RD Costa E Ortiacute HJ Bolink S Graber S Schaffner M Neuburger CE Housecroft EC Constable Adv Funct Mater 19 3456 (2009)
18 F De Angelis F Santoro MK Nazeruddin V Barone J Phys Chem B 112 13181(2008)
19 HW Ham YS Kim Mol Cryst Liquid Cryst 520 384 (2010)20 T Li M Cui G Ran Q Song Dyes Pigments 112 1 (2015)21 X Qu Y Liu G Godefroid Y Si X Shang X Wu Z Wu Eur J Inorg Chem
3370 (2013)22 X Qu Y Liu Y Si X Wu Z Wu Dalton Trans 43 1246 (2014)23 KPS Zanoni BK Kariyazaki AA Ito MK Brennaman TJ Meyer NYM Iha Inorg
Chem 53 4089 (2014)24 RD Costa E Ortiacute D Tordera A Pertegaacutes HJ Bolink S Graber CE Housecroft L
Sachno M Neuburger EC Constable Adv Energy Mater 1 282 (2011)25 L Sun A Galan S Ladouceur JD Slinker E Zysman-Colman J Mater Chem 21
18083 (2011)26 MdK Nazeeruddin RT Wegh Z Zhou C Klein Q Wang F De Angelis S Fantacci M
Graumltzel Inorg Chem 45 9245 (2006)27 S Sprouse KA King PJ Spellane RJ Watts J Am Chem Soc 106 6647 (1984)28 CD Ertl J Cerda JM Junquera-Hernaacutendez A Pertegaacutes HJ Bolink EC Constable M
Neuburger E Ortiacute CE Housecroft RSC Adv 5 42815 (2015)29 D Tordera M Delgado E Ortiacute HJ Bolink J Frey MK Nazeeruddin E Baranoff
Chem Mater 24 1896 (2012)30 KA McGee KR Mann Inorg Chem 46 7800 (2007)31 F Monti A Baschieri I Gualandi JJ Serrano-Perez JM Junquera-Hernaacutendez D
Tonelli A Mazzanti S Muzzioli S Stagni C Roldan-Carmona A Pertegaacutes HJ BolinkE Ortiacute L Sambri N Armaroli Inorg Chem 53 7709 (2014)
32 D Ma L Duan Y Qiu Dalton Trans 44 8521 (2015)33 GE Schneider HJ Bolink EC Constable CD Ertl CE Housecroft A Pertegaacutes JA
Zampese A Kanitz F Kessler SB Meier Dalton Trans 43 1961 (2014)34 B Schmid FO Garces RJ Watts Inorg Chem 33 9 (1994)35 AM Buumlnzli EC Constable CE Housecroft A Prescimone JA Zampese G Longo L
Gil-Escrig A Pertegaacutes E Ortiacute HJ Bolink Chem Sci 6 2843 (2015)36 Y Shen DD Kuddes CA Naquin TW Hesterberg C Kusmierz BJ Holliday JD
Slinker Appl Phys Lett 102 article 203305 (2013)37 KJ Suhr LD Bastatas Y Shen LA Mitchell BJ Holliday JD Slinker ACS Appl
Mater Interfaces 8 8888 (2016)38 HJ Bolink L Cappelli E Coronado M Graumltzel E Ortiacute RD Costa PMA Viruela
MdK Nazeeruddin J Am Chem Soc 128 14786 (2006)39 DB Moon YS Choe Mol Cryst Liq Cryst 584 60 (2013)40 X Zeng M Tavasli IFA Perepichka AS Batsanov MR Bryce C-J Chiang C Rothe
AP Monkman Chem Eur J 14 933 (2008)41 C Rothe C-J Chiang V Jankus K Abdullah X Zeng R Jitchati AS Batsanov MR
Bryce AP Monkman Adv Funct Mater 19 2038 (2009)42 H-F Chen K-T Wong Y-H Liu YA Wang Y-M Cheng M-W Chung P-T Chou
H-C Su J Mater Chem 21 768 (2011)43 D Tordera A Pertegaacutes NMA Shavaleev R Scopelliti E Ortiacute HJ Bolink E Baranoff
M Graumltzel MK Nazeeruddin J Mater Chem 22 19264 (2012)44 AB Tamayo BD Alleyne PI Djurovich S Lamansky I Tsyba NN Ho R Bau ME
Thompson J Am Chem Soc 125 7377 (2003)
200 CE Housecroft and EC Constable
45 AB Tamayo S Garon T Sajoto PI Djurovich IM Tsyba R Bau ME ThompsonInorg Chem 44 8723 (2005)
46 CD Sunesh M Chandran G Mathai Y Choe Opt Mater 35 407 (2013)47 CD Sunesh S Ok G Mathai Y Choe Thin Solid Films 531 530 (2013)48 See for example M Felici P Contreras-Carballada JMM Smits RJM Nolte RM
Williams L De Cola MC Feiters Molecules 15 2039 (2010)49 S Ladouceur D Fortin E Zysman-Colman Inorg Chem 50 11514 (2011)50 L He Z Wang C Yang L Duan R Tang X Song C Pan Dyes Pigments 131 76
(2016)51 L He L Duan J Qiao R Wang PA Wei L Wang Y Qiu Adv Funct Mater 18 2123
(2008)52 T Hu L Duan J Qiao L He D Zhang R Wang L Wang Y Qiu Org Electr 13 1948
(2012)53 CD Sunesh MS Subeesh K Shanmugasundaram RK Chitumalla J Jang Y Choe
Dyes Pigments 128 190 (2016)54 CD Sunesh K Shanmugasundaram MS Subeesh RK Chitumalla J Jang Y Choe A
CS Appl Mater Interfaces 7 7741 (2015)55 CD Sunesh G Mathai Y Choe Org Electron 15 667 (2014)56 L He J Qiao L Duan G Dong D Zhang L Wang Y Qiu Adv Funct Mater 19 2950
(2009)57 CD Sunesh G Mathai Y Choe ACS Appl Mater Interfaces 6 17416 (2014)58 M Mydlak C Bizzarri D Hartmann W Sarfert G Schmid L De Cola Adv Funct
Mater 20 1812 (2010)59 Y Kwon Y Choe J Solut Chem 43 1710 (2014)60 H-C Su H-F Chen F-C Fang C-C Liu C-C Wu K-T Wong Y-H Liu S-M Peng
J Am Chem Soc 130 3413 (2008)61 L He J Qiao L Duan G Dong D Zhang L Wang Y Qiu Adv Funct Mater 19 2950
(2009)62 J Zhang L Zhou HA Al-Attar K Shao L Wang D Zhu Z Su MR Bryce A
P Monkman Adv Funct Mater 23 4667 (2013)63 EC Constable CE Housecroft GE Schneider JA Zampese HJ Bolink A Pertegaacutes C
Roldan-Carmona Dalton Trans 43 4653 (2014)64 DP Ris GE Schneider CD Ertl E Kohler T Muumlntener M Neuburger EC Constable
CE Housecroft J Organomet Chem 812 272 (2016)65 JL Rodriguez-Redondo RD Costa E Ortiacute A Sastre-Santos HJ Bolink F
Fernaacutendez-Laacutezaro Dalton Trans 9787 (2009)66 CD Ertl PhD Thesis University of Basel (2016)67 S Graber K Doyle M Neuburger CE Housecroft EC Constable RD Costa E Ortiacute
D Repetto HJ Bolink J Am Chem Soc 130 14944 (2008)68 HJ Bolink E Coronado RD Costa E Ortiacute M Sessolo S Graber KA Doyle M
Neuburger CE Housecroft EC Constable Adv Mater 20 3910 (2008)69 RD Costa E Ortiacute HJ Bolink S Graber CE Housecroft EC Constable Chem
Commun 47 3207 (2011)70 RD Costa E Ortiacute HJ Bolink S Graber CE Housecroft M Neuburger S Schaffner E
C Constable Chem Commun 2029 (2009)71 RD Costa E Ortiacute HJ Bolink S Graber CE Housecroft EC Constable Adv Funct
Mater 20 1511 (2010)72 D Tordera S Meier M Lenes RD Costa E Ortiacute W Sarfert HJ Bolink Adv Mater
24 897 (2012)73 NM Shavaleev R Scopelliti M Graumltzel MK Nazeeruddin A Pertegaacutes C
Roldan-Carmona D Tordera HJ Bolink J Mater Chem C 1 2241 (2013)74 AM Buumlnzli HJ Bolink EC Constable CE Housecroft M Neuburger E Ortiacute A
Pertegaacutes JA Zampese Eur J Inorg Chem 3780 (2012)
7 Development of Cyclometallated Iridium(III) Complexes hellip 201
75 GE Schneider A Pertegaacutes EC Constable CE Housecroft N Hostettler CD Morris JA Zampese HJ Bolink JM Junquera-Hernaacutendez E Ortiacute M Sessolo J Mater ChemC 2 7047 (2014)
76 P Li G-G Shan H-T Cao D-X Zhu Z-M Su R Jitchati MR Bryce Eur J InorgChem 2376 (2014)
77 RD Costa E Ortiacute HJ Bolink S Graber CE Housecroft EC Constable J Am ChemSoc 132 5978 (2010)
78 L He L Duan J Qiao D Zhang L Wang Y Qiu Chem Commun 47 6467 (2011)79 H-F Chen W-Y Hung S-W Chen T-C Wang S-W Lin S-H Chou C-T Liao H-
C Su H-A Pan P-T Chou Y-H Liu K-T Wong Inorg Chem 51 12114 (2012)80 EC Constable Chem Soc Rev 36 246 (2007)81 D Tordera JJ Serrano-Peacuterez A Pertegaacutes E Ortiacute HJ Bolink E Baranoff MdK
Nazeeruddin J Frey Chem Mater 25 3391 (2013)82 CD Sunesh G Mathai Y-R Cho Y Choe Polyhedron 57 77 (2013)83 Y Kwon CD Sunesh Y Choe Opt Mater 39 40 (2015)84 S Cha Y Choe Mol Cryst Liq Cryst 601 205 (2014)85 NM Shavaleev G Xie S Varghese DB Cordes AMZ Slawin C Momblona E Ortiacute
HJ Bolink IDW Samuel E Zysman-Colman Inorg Chem 54 5907 (2015)86 D Tordera AM Buumlnzli A Pertegaacutes JM Junquera-Hernaacutendez EC Constable JA
Zampese CE Housecroft E Ortiacute HJ Bolink Chem Eur J 19 8597 (2013)87 CD Ertl L Gil-Escrig J Cerdaacute A Pertegaacutes HJ Bolink JM Junquera-Hernaacutendez A
Prescimone M Neuburger EC Constable E Ortiacute CE Housecroft Dalton Trans 4511668 (2016)
88 S Evariste M Sandroni TW Rees C Roldan-Carmona L Gil-Escrig HJ Bolink EBaranoff E Zysman-Colman J Mater Chem C 2 5793 (2014)
89 C-H Yang J Beltran V Lemaur J Cornil D Hartmann W Sarfert RA Froumlhlich CBizzarri L De Cola Inorg Chem 49 9891 (2010)
90 F Kessler RD Costa D Di Censo R Scopelliti E Ortiacute HJ Bolink S Meier W SarfertM Graumltzel MdK Nazeeruddin E Baranoff Dalton Trans 41 180 (2012)
91 SB Meier W Sarfert JM Junquera-Hernaacutendez M Delgado D Tordera E Ortiacute HJ Bolink F Kessler R Scopelliti M Graumltzel MK Nazeeruddin E Baranoff J MaterChem C 1 58 (2013)
92 F Zhang L Duan J Qiao G Dong L Wang Y Qiu Org Electron 13 2442 (2012)93 RD Costa R Casillas J Cano J Phys Chem C 117 8545 (2013)94 HJ Bolink L Cappelli E Coronado A Parham P Stoumlssel Chem Mater 18 2778 (2006)95 E Zysman-Colman JD Slinker JB Parker GG Malliaras S Bernhard Chem Mater 20
388 (2008)96 H-C Su H-F Chen C-C Wu K-T Wong Chem Asian J 3 1922 (2008)97 T-H Kwon YH Oh I-S Shin J-I Hong Adv Funct Mater 19 711 (2009)98 H-F Chen C Wu M-C Kuo ME Thompson K-T Wong J Mater Chem 22 9556
(2012)99 V Florini A DrsquoIgnazio KDM Magee MI Ogden M Massi S Stagni Dalton Trans
45 3256 (2016)100 V Florini S Zacchini P Raiteri R Mazzoni V Zanotti M Massi S Stagni Dalton
Trans 45 12884 (2016)101 IJ Bruno JC Cole PR Edgington M Kessler CF Macrae P McCabe J Pearson
R Taylor Acta Cryst B 58 389 (2002)102 CF Macrae IJ Bruno JA Chisholm PR Edgington P McCabe E Pidcock
L Rodriguez-Monge R Taylor J van de Streek PA Wood J Appl Cryst 41 466 (2008)
202 CE Housecroft and EC Constable
Chapter 8Recent Advances on Blue-EmittingIridium(III) Complexes for Light-EmittingElectrochemical Cells
Lei He
Abstract Blue light-emitting electrochemical cells (LECs) are critically importantfor the fabrication of white LECs toward solid-state lighting applications Duringthe past decade blue-emitting iridium(III) complexes have been pursued as theemitting materials for blue LECs In this chapter the recent advances onblue-emitting iridium(III) complexes for LECs are discussed based on the fourmolecular design strategies namely (i) modification on the archetype complex [Ir(pp)2(bpy)]
+ (Hppy is 2-phenylpyridine bpy is 22rsquo-bipyridine) (ii) using ancillaryligands beyond the bpy skeleton (iii) using ancillary ligands with strong ligandfield strength and (iv) using cyclometalating ligands beyond the ppy skeleton Thediscussion is focused on how to blue-shift the emission of the complexes as well asthe performances of the complexes in LECs This complements Chap 7 in whichthe design of iridium (III) complexes for stable efficient and fast LECs isdescribed So far near-UV emitting complexes and LECs have been reportedNevertheless good color purity high efficiency high brightness and short responsetime have not been achieved in one blue LEC Most importantly the stability of theblue LECs should be significantly improved With high-performance blue-emittingmaterials and optimized devices it is believed that blue LECs suitable for appli-cations will come out in the near future
Keywords Light-emitting Electrochemical cells Thin-film devices Blueemission Iridium(III) complexes Ligand optimization
L He (amp)Department of Chemistry Central China Normal University WuhanHubei Province 430079 Peoplersquos Republic of Chinae-mail heleiccnu126com helei06csueducn
L HeCollege of Chemistry and Chemical Engineering Central South UniversityChangsha Hunan Province 410083 Peoplersquos Republic of China
copy Springer International Publishing AG 2017RD Costa (ed) Light-Emitting Electrochemical CellsDOI 101007978-3-319-58613-7_8
203
81 Introduction
With intrinsic ionic nature luminescent ionic iridium(III) complexes have arousedsignificant research interest in the past decade for their application as emittingmaterials in solid-state light-emitting electrochemical cells (LECs) [1ndash4] In LECsthe complexes harvest both singlet and triplet excitons leading to high electrolu-minescent efficiencies More importantly they bear largely destabilizednon-radiative metal-centered states (3MC) and are thus suitable for the achievementof efficient blue light emission [1] With these merits ionic iridium(III) complexeshold the promise for attaining highly efficient LECs with the emission colorsspanning the entire visible spectrum [3 4]
Throughout the development of LECs incorporating ionic iridium(III) com-plexes a clear roadmap is to develop blue-emitting complexes and devices withhigh color purity high efficiency and enhanced stability Huge efforts have beendevoted since the blue light is critically important for the generation of white lighttoward solid-state lighting applications Nowadays orange-red LECs with highefficiency and stability have already been achieved [4 5] and green LECs with highefficiency and improved stability are on the way [6] Nevertheless the overallperformances of blue LECs including color purity efficiency and stability lag farbehind those of red and green LECs which has emerged as a serious bottleneck forwhite LECs For making blue devices high-performance blue-emitting materialsare the first prerequisite This chapter discusses the recent advances onblue-emitting ionic iridium(III) complexes for LECs with an outlook in this area atthe end The discussion is focused on how to blue-shift the emission of the com-plexes for the application in LECs Other blue-emitting complexes which havebeen developed without applications in LECs will not be focused here Thiscomplements Chap 7 in which the design of iridium(III) complexes for stableefficient and fast LECs is described Finally although the definition of the LECfigures-of-merit is provided the reader will find more details in Chap 1
82 Blue-Emitting Iridium(III) Complexes for LECs
In most cases ionic iridium(III) complexes have a form of [Ir(C^N)2(N^N)]+Aminus
where C^N is the anionic cyclometalating ligand N^N is the neutral ancillary ligandand Aminus is the counter anion As shown in Chap 7 the archetype complex used forLECs is [Ir(ppy)2(bpy)][PF6] (complex 1 Fig 81) [7ndash10] where Hppy is2-phenylpyridine (C^N) bpy is 22rsquo-bipyridine (N^N) and PF6
minus is hexafluorophos-phate (Aminus) In 2004 Slinker and co-authors for the first time applied this type ofcomplexes for LECs [1] The complex they used was [Ir(ppy)2(dtb-bpy)][PF6] (2)(dtb-bpy is 44rsquo-di-tert-butyl-22rsquo-bipyridine) an analog of complex 1 With itshighest occupied molecular orbital (HOMO) and lowest unoccupied molecular
204 L He
orbital (LUMO) levels at minus56 and minus30 eV respectively complex 1 has an energygap of 24 eV and emits orange-red light peaked around 590 nm [7 9] The emissionfrom complex 1 has predominant ligand-to-ligand charge-transfer (LLCT) (ppybpy) and metal-to-ligand charge-transfer (MLCT) (Ir bpy) charactermdashseeChap 7 for more details [9] LECs incorporating complex 1 emit orange-red lightwith the peak current efficiency boosted to 156 cdA [9] Since the Slinkerrsquos workresearchers have devoted tremendous efforts to pushing the emission of the complextoward blue To this end four strategies have been followed namely (i) modificationon [Ir(ppy)2(bpy)]
+ (ii) using ancillary ligands beyond the bpy skeleton (iii) usingancillary ligands with strong ligand field strength and (iv) using cyclometalatingligands beyond the ppy skeleton In this chapter the advances on the blue-emittingionic iridium(III) complexes are elaborated on the basis of the four strategies
821 Modification on [Ir(ppy)2(bpy)]+
To blue-shift the emission of the complex a straightforward approach is to widenthe energy gap between the HOMO and LUMO because a widened energy gapusually leads to blue-shifted emission It has been established that the HOMO of [Ir(ppy)2(bpy)]
+ resides on the iridium ion and the phenyl rings of ppy and theLUMO is delocalized over bpy [9 10] The orthogonal distributions of the frontiermolecular orbitals allow independent tunings on the HOMO and LUMO levelsthrough chemically modifying the ppy or bpy ligands In general enlarging theenergy gap of [Ir(ppy)2(bpy)]
+ can be done with two approaches (i) stabilizing the
3
65
N
IrN
NN
+
F
F
N
IrN
NN
+
F
F
F
F
N
IrN
NN
+
F
F
F
F
4
N
IrN
NN
+
CF3
CF3
F
F
F
F
N
IrN
NN
[PF6]-+
R=H 1 R=tBu 2
R
R
[PF6]- [PF6]-
[PF6]-[PF6]-
Fig 81 Ionic iridium(III) complexes (except 1) developed from modifications on [Ir(ppy)2(bpy)]
+
8 Recent Advances on Blue-Emitting Iridium(III) Complexes hellip 205
HOMO via attaching electron-withdrawing groups to the phenyl rings of ppy and(ii) destabilizing the LUMO via attaching electron-donating groups to bpy All thecomplexes discussed in this section are displayed in Figs 81 and 82
By substituting the phenyl ring of each ppy-type C^N ligand with anelectron-withdrawing fluorine atom Slinker and co-authors developed greenemitting complex 3 of which the light emission was blue-shifted for about 45 nmrelative to that of the parent complex 2 due to the stabilization of HOMO inducedby the fluorine substitution [11] Lowry and co-authors developed a series ofcomplexes with fluorinated ppy ligands and found that the emission was stepwiselyblue-shifted upon adding more fluorine atoms to the phenyl rings of ppy due to thegradual stabilization of the HOMO [12] Among the complexes they developedcomplex 4 gave the mostly blue-shifted emission in solution with the majoremission peak at 470 nm owing to the additional electron-withdrawing minusCF3group attached to the C^N ligand [12] The LEC ITO4 (70 minus 80 nm)Au howeverfeatured red-shifted emission centered at 500 nm owing to the strong inter-molecular interactions in the emissive layer Under minus60 V (ITO biased negatively)this green-blue LEC provided a peak external quantum efficiency (EQE) of 016and a maximum brightness of 33 cdm2
Bolink and co-authors developed a blue-green emitting complex 5 by usingfluorine substituted ppy 2-(24-difluorophenyl)pyridine (dfppy) as the C^N ligandand methyl-substituted 110-phenanthroline (phen) as the N^N ligand [13] Thephen ligand is analogous to bpy and the complex [Ir(ppy)2(phen)][PF6] bearsenergy levels and emission energy quite similar to those of complex 1 [14] Inacetonitrile solution complex 5 emitted blue-green light peaked at 476 and 508 nmblue-shifted by over 50 nm relative to that of the un-fluorinated counterpart [14]However in the concentrated films complex 5 showed largely red-shifted emissionbecause the strong intermolecular interactions made the charge-transfer states shiftdownward to become the lowest emitting states Due to this red-shift the LEC ITO5Al emitted green-yellow light centered around 560 nm
Fluorine substitution to the phenyl rings of the C^N ligands has been widelyadopted for blue-shifting the emission of ionic iridium(III) complexes Thisapproach has been proven quite effective to stabilize the HOMO without largelyaltering the LUMO for a net enlargement of the energy gap However it appearsthat fluorine substitution tends to lower the brightness of LECs By fluorinating theppy ligands of complex 2 Bolink and co-authors developed complex 6 whichemitted green light centered around 512 nm in solution [15] The green LECITOPEDOTPSS (100 nm)6 [BMIM][PF6] (molar ratio 111) (100 minus 200 nm)Al ([BMIM][PF6] is 1-butyl-3-methylimidazolium hexafluorophosphate) showed apeak EQE of 149 under 30 V which represents the record value for LECsincorporating iridium(III) complexes Nevertheless the maximum brightness of theLEC (ca 20 cdm2 at 30 V) is one order of magnitude lower than that (330 cdm2
at 3 V) of ITO2Au [1] even though complex 6 phosphoresces more efficientlythan complex 2 and the latter device is not doped with ionic liquid and uses Au as amore inert cathode Similar conclusions can be drawn between complexes 2 and 3[1 11] It appears that fluorine substitution makes the carrier injection andor
206 L He
transport in the emissive layer more difficult lowering the the current flux andbrightness of the device This issue has always been encountered for LECs incor-porating iridium(III) complexes with enlarged energy gaps One reason could bethat the deep HOMO or shallow LUMO associated with the enlarged energy gapresults in more difficult injection of holes or electrons from the inert electrodes
Fluorine substitution to the C^N ligands has been found to deteriorate the thermaland electrochemical stability of the complex due to the highly polarized and reactiveC-F bonds This issue has been elucidated in neutral blue-emitting iridium(III)complexes targeted for OLEDs [16] Tordera et al and co-authors developed a seriesof complexes with the C^N ligands substituted with fluorine atoms of variednumbers and positions [6 17] They demonstrated clearly that increasing the fluorinecontent of the complexes decreased the stability of LECs [17]
To blue-shift the emission and meanwhile maintain the stability of the complexesother electron-withdrawing substituents rather than fluorine have been testedTordera et al and Constable et al utilized ndashSO2R as the electron-withdrawinggroups for the complexes [18 19] With the sulfone-containing ppy as the C^Nligand and bpy as the N^N ligand Constable et al developed complexes 7ndash9 [19]which gave green-blue emission peaked around 493 and 525 nm in solutionblue-shifted for about 90 nm than that of complex 1 It was shown that ndashSO2Rsubstitution stabilized the HOMO levels of the complexes with an extent similar tothat caused by fluorine substitution Ertl and co-authors further developed complexes10ndash11 and investigated the effect of the substitution position of minusSO2Me group onthe emission properties of the complexes [20] In acetonitrile solution complexes
R=CH3 12 R=tbu 13
R1= -OCH3 R2= -CH3 R3= -CH3 R4=H 16R1=R2= -OCH3 R3= -CH3 R4=H 17
R1= -CH3 R2=R3= H R4=-CH3 19R1=R2= -OCH3 R2= R4=H 18
R=H 20 R=F 21
N
IrN
NN
+
R
R
F3C
F3C
NN
Ir
N
N
NN
+
R1
R1
R2
R2
R4
R4
R3
R3
N
IrN
NN
+
R
R
R
R
N
NN
IrN
NN
+
F5S
F5S
34
5
4-SF5 14 5-SF5 15
R=CH3 7 R=tBu 8 R=nC12H25 9 4-MeO2S 10 3-MeO2S 11
N
IrN
NN
+
RO2S
RO2S
N
IrN
NN
+
MeO2S
MeO2S
34
5
[PF6]-
[PF6]-
[PF6]- [PF6]-
[PF6]-[PF6]-
Fig 82 Ionic iridium(III) complexes developed from modifications on [Ir(ppy)2(bpy)]+
8 Recent Advances on Blue-Emitting Iridium(III) Complexes hellip 207
10ndash11 emitted highly efficient green-blue light with the emission maximum at 493ndash517 nm Placing the minusSO2Me group para to the Ir-C bond stabilized the HOMOmore than that at the meta position In the solid state the emission from the com-plexes was red-shifted to the green-yellow region due to the strong intermolecularinteractions For this reason LECs based on the complexes gave yellow emissionpeaked at 550ndash560 nm
As a strongly electron-withdrawing group trifluoromethyl (minusCF3) has beenemployed for blue-shifting the emission of the complexes Shavaleev andco-authors developed complexes 12ndash13 with the minusCF3 substituted ppy ligands [21]Substitution of minusCF3 at the position para to the Ir-C bond stabilized the HOMO by03 eV In solution complexes 12-13 emitted green light peaked around 510 nmThe LEC ITOPEDOTPSS (80 nm)12 or 13 [BMIM][PF6] (molar ratio 1 025)(100 nm)Al however showed yellow-green electroluminescence (EL) peakedaround 547 nm This red-shift between the PL and EL has often been observed forLECs using blue iridium(III) emitters which is in general attributed to the strongintermolecular interactions in the emissive layers andor the chemical degradationof the complexes under electrical excitation as will be discussed later
Shavaleev and co-authors used a chemically inert very strong electron-withdrawing group sulfur pentafluoride (minusSF5) for blue-emitting complexes [22]Substitution of ndashSF5 on the ppy stabilized the HOMO by 037 eV almost inde-pendent of the substitution position (para or meta to the Ir-C bond) By attachingminusSF5 to the ppy ligands in complex 2 complexes 14-15 were developed whichgave very efficient green phosphorescence centered at 482ndash519 nm Neverthelessintroduction of minusSF6 resulted in irreversible electrochemical reduction of thecomplexes For this reason LECs made with the complexes were non-emissive
The pyridine is an intrinsically electron-deficient aromatic ring and has beenused to replace the phenyl ring in ppy for fluorine-free iridium(III) emitters Usingmethoxy- and methyl-substituted 23rsquo-bipyridine as the C^N ligands Evariste andco-authors developed fluorine-free ionic iridium(III) emitters 16-19 [23] whichgave green-blue emission centered around 515 nm in solution blue-shifted formore than 70 nm relative to that of complex 2 However complexes 16ndash19 gavecomplicated irreversible oxidation processes in solution which would negativelyinfluence their performances in LECs The methoxyl substitution near the nitrogenatom in the pyridine hardly influenced the emission wavelength but helped toimpede the aggregation of the complexes in the solid-state The LECs ITOPEDOTPSS (80 nm)16ndash19 [BMIM][PF6] (1025) (100 nm)Al emitted green light withthe emission peaks and shoulders at around 550 and 500 nm respectively [23] Thered-shift between the solutions PL and EL was caused by the strong intermolecularinteractions in the emissive layers Similar to the fluorinated complexesfluorine-free complexes 16ndash19 showed limited stability in LECs Thereforeavoiding fluorine substitution on the complexes is not adequate for improving thestability of LECs
208 L He
So far stabilizing the HOMO of [Ir(ppy)2(bpy)]+ via attaching electron-
withdrawing groups has only resulted in light emission in the green or green-blueregion To further enlarge the energy gap and blue-shift the emission of the com-plex the LUMO needs to be destabilized A straightforward approach is to attachelectron-donating groups at the bpy ligand where the LUMO is placed Byattaching the strongly electron-donating dimethylamino group at the bpy in com-plex 1 Nazeeruddin and co-authors developed complex 20 [24] The dimethy-lamino substitution destabilized the LUMO by 03 eV with a slight destabilizationof the HOMO by 01 eV leading to a net enlargement of the energy gap In ace-tonitrile solution complex 20 emitted green-blue light with the emission maximumand shoulder at 520 and 491 nm respectively blue-shifted by 70 nm relative to theemission of complex 1 The LEC ITO20 (100 nm)Ag gave green-blue lightsimilar to the PL in the film or solution with a maximum brightness of 200 cdm2 acurrent efficiency of 043 cdA and an EQE of 02 under 35 V [24] By fluori-nating the ppy ligands of complex 20 Angelis and co-authors further developedcomplex 21 [25] Owing to the destabilized LUMO and the stabilized HOMOcomplex 21 had a large energy gap of 313 eV and emitted blue-green light inacetonitrile solution with the emission maximum at 493 nm and the emissionshoulder at 463 nm No LECs based on complex 21 were reported
822 Using Ancillary Ligands Beyond the bpy Skeleton
For phosphorescent ionic iridium(III) complexes bpy-type ligands are the mostwidely employed ancillary ligands However the LUMO levels associated with thebpy-type ligands are usually low due to the electron-deficient character of thepyridine rings Although substituting bpy with electron-donating groups destabi-lizes the LUMO levels the overall blue-shifting effect on the emission of thecomplex is limited leading to the major emission peak located around 490 nm evenwhen fluorine substitutions are presented on the C^N ligands (complex 21) Todevelop truly-blue ionic iridium(III) complexes other ancillary ligands beyond thebpy skeleton are highly desired All the complexes described in this section aredisplayed in Figs 83 and 84
He and co-authors proposed that inserting electron-donating atoms such asnitrogen into the N^N ligand would remarkably increase the electron density on theligand destabilizing the LUMO level of the complex [26] To prove this designthey synthesized 2-(1H-pyrazol-1-yl)pyridine (pzpy) as the N^N ligand to preparethe blue-emitting complexes 22 and 23 [26] The nitrogen atom connected to thepyridine ring in pzpy bears a lone pair of electrons which are conjugated into thepyridine-pyrazole plane rendering pzpy with high electron density Compared tocomplex 1 complex 22 showed a significantly destabilized (04 eV) LUMO leveland a nearly undisturbed HOMO level In acetonitrile solution complex 22 emittedblue-green light peaked at 475 and 505 nm blue-shifted by over 100 nm relative tothat of complex 1 The experimental results clearly demonstrated the feasibility of
8 Recent Advances on Blue-Emitting Iridium(III) Complexes hellip 209
the proposed design strategy For complexes 1 and 22 their emitting triplet statesare completely different in nature ie the lowest emitting triplet state in complex22 is a mixture of ppy-centered 3p minus p and 3MLCT (Ir ppy) while in complex1 it is mixed 3LLCT (ppy bpy) and 3MLCT (Ir bpy) With the dfppy C^Nligand complex 23 gave further blue-shifted emission with the emission maximumat 452 nm [26] In neat films the emission from complexes 22 and 23 exhibitedonly small red-shifts compared to their PL spectra in solution The LECsITOPEDOTPSS (50 nm)22 or 23 (75 nm)Al gave blue-green light peaked at486 nm for complex 22 and blue light peaked at 460 nm for complex 23 [26] TheLECs showed maximum brightness and current efficiency of 52 cdm2 096 cdAfor complex 22 under 50 V and 39 cdm2 065 cdA for complex 23 under 65 VLECs were also fabricated with doped ionic liquid ITO22 or 23 [BMIM][PF6](molar ratio 1 035) (120 nm)Al The blue-green LEC incorporating complex 22and the ionic liquid showed a maximum brightness of 94 cdm2 and a peak currentefficiency of 43 cdA under 50 V which were largely increased compared to thepristine device without ionic liquid However for the LEC incorporating complex23 and the ionic liquid the EL was red-shifted to the yellow region This red-shiftwas presumably caused by the polarized molecular orbitals or chemical degradationof the complexes under electrical excitation It was noted that the LECs withcomplex 22 showed much lower performances than the LECs with complex 1 [9]which could be attributed to (i) the larger electron-injection barrier at thecathodeemissive layer interface for the former devices caused by the high LUMO
N
IrN
N
+
R=H 22 R=F 23
R
R
R
R
NN
N
IrN
N
+
R
R
R=H 24 R=F 25
NN
Ph
R
R
N
IrN
N
+
R1=R2=H R3=Me 26 R1=H R2=R3=Me 27
R1
R1
R1
R1
NN
R3
R2
R1=F R2=H R3=Me 28 R1=F R2=R3=Me 29
N
IrN
N
+
5-SO2Me R=H 30 5-SO2Me R=Me 314-SO2Me R=H 32 4-SO2Me R=Me 33
NN
R
R
MeO2S
MeO2S
34
56
N
IrN
N
+
34
NN
Ph
[PF6]- [PF6]- [PF6]-
[PF6]-[PF6]-
Fig 83 Ionic iridium(III) complexes using ancillary ligands beyond the bpy skeleton
210 L He
level of complex 22 and (ii) impeded electron transport in the emissive layers forthe former devices as indicated by the irreversible electrochemical reductionoccurring on pzpy in complex 22 [26]
Since the report by He et al the pzpy-type ancillary ligands have been widelyemployed for constructing blue-emitting complexes For these complexes theemission usually comes from the triplet states composed of C^N-centered 3p minus p
and 3MLCT (Ir C^N) ie the emission is mainly determined by the C^Nligands Sunesh and co-authors modified the pzpy ligand by anchoring a phenylring next to the coordinated nitrogen atom in the pyrazole and developed complexes24 and 25 [27] which showed emission and electrochemical properties quite similarto those of complexes 22 and 23 Nevertheless they showed lower luminescentefficiencies than complexes 22 and 23 owing to the phosphorescence quenchingcaused by the steric hindrance of the pendent phenyl ring on the ancillary ligand[28] The LEC ITOPEDOTPSS (50 nm)24 or 25 (75 nm)Al emitted blue-greenand blue light for complexes 24 and 25 respectively [27] Driven with sweepingvoltages at 05 V sminus1 the LECs gave peak current efficiencies of 046 and038 cdA for complexes 24 and 25 respectively
Sunesh and co-authors further modified the pzpy ligand by attaching methylgroups on the pyrazole and developed complexes 26ndash29 with the ppy or dfppyC^N ligands [29] In solution complexes 26 and 27 gave blue-green emissionwhereas complexes 28 and 29 gave blue emission Complexes 26ndash29 showed lowerluminescent efficiencies than their parent complexes 22 and 23 due to the sterichindrance of the methyl groups ortho to the Ir-N bonds [28] The LECsITOPEDOTPSS26ndash29Al gave yellow light for complexes 26ndash27 and green lightcentered around 505 nm for complexes 28ndash29 [29] The EL was largely red-shiftedcompared to the PL spectra in solution which was tentatively attributed to thepolarization effect under electrical excitation Driven with sweeping voltages at5 V sminus1 the LECs showed current efficiencies below 04 cdA
With pzpy or methyl-substituted pzpy as the N^N ligands and sulfonyl-substitutedppy as the C^N ligands Ertl and co-authors developed fluorine-free blue-emittingcomplexes 30ndash33 with varied substitution positions of the sulfonyl group on thephenyl rings of the C^N ligands [30] In acetonitrile solution complexes 30 and 31emitted green-blue light peaked around 491 and 523 nm while complexes 32 and 33emitted blue light peaked around 463 and 493 nm In the solid-state the complexesshowed red-shifted (ca 30 nm) emission with largely decreased luminescent effi-ciencies owing to the strong intermolecular interactions The LECs ITOPEDOTPSS (80 nm)30ndash33 [BMIM][PF6] (molar ratio 1 025) (100 nm)Al gave greenemission peaked around 500 and 540 nm [30] The red-shift of the EL compared tothe PL in solution was attributed to the strong intermolecular interactions in thesolid-state Driven under a pulsed current of 100 Amminus2 the LECs showed maximumbrightness below 141 cdm2 and peak EQEs below 04 Although fluorine-free andcontaining hydrophobic methyl substitutes complexes 30ndash33 still showed limitedstability in the LECs
By changing the position of the electron-donating nitrogen atom He andco-authors developed another pyridine-pyrazole ancillary ligand ie 2-
8 Recent Advances on Blue-Emitting Iridium(III) Complexes hellip 211
(1-phenyl-1H-pyrazol-3-yl)pyridine (phpzpy) and assembled a blue-green-emittingcomplex 34 with the ppy C^N ligand [31] Through the nitrogen atom in thepyrazole of phpzpy a pendant phenyl ring was attached ortho to the Ir-N bond Thephpzpy ligand brought about two advantages (i) blue-shifting the emission of thecomplex in a manner similar to pzpy and (ii) enhancing the intrinsic stability of thecomplex through the intermolecular p-p stacking formed between the pendantphenyl ring in phpzpy and the phenyl ring of one ppy ligand [5] Moreovercomplex 34 showed a reversible electrochemical reduction occurring on phpzpywhich indicated enhanced electrochemical stability and facilitated electron transportfor complex 34 in LECs In acetonitrile solution complex 34 emitted blue-greenlight peaked at 480 nm with remarkable ppy-centered 3p minus p character In thediluted solid matrix the efficiency of complex 34 was significantly enhanced rel-ative to that in solution due to the the restricted rotation of the pendant phenyl ringin phpzpy The LEC ITOPEDOTPSS (40 nm)34 [BMIM][PF6] (molar ratio105) (100 nm)Al gave blue-green light peaked at 486 nm [31] Driven under30 V the LEC showed a maximum brightness of 37 cdm2 a peak current effi-ciency of 87 cdA and a half lifetime of nearly 10 h The overall performancesespecially the stability were remarkably enhanced compared to the LEC incorpo-rating complex 22
Following their molecular design strategy ie inserting electron-donatingnitrogen atoms into the ancillary ligand He and co-authors further developedpyridine-imidazole ancillary ligands and assembled complexes 35ndash36 for LECs[32] With 2-(1-phenyl-1H-imidazol-2-yl)pyridine (pyim) as the N^N ligand andppy as the C^N ligand complex 35 emitted yellow light centered at 550 nmblue-shifted by about 40 nm relative to that of complex 1 [32] This blue-shift wassmaller than that (gt 100 nm) observed for complex 22 consistent with the smallerLUMO destabilization (02 eV) caused by pyim than that (gt 04 eV) caused bypzpy The emission from complex 35 exhibited dominant 3LLCT (ppy pyim)and 3MLCT (Ir pyim) character in contrast to the dominant ppy-centered 3p-p
character for complex 22 Therefore both the extent of the blue-shifting effect andthe nature of the emissive triplet state depend on the relative positions of theelectron-donating nitrogen atoms in the five-membered N-heterocycle of theancillary ligand [33] With the dfppy C^N ligand complex 36 showed furtherblue-shifted emission with the emission peak at 489 nm in acetonitrile solution[32] Moreover the complexes with pyridine-imidazole ancillary ligands showedreversible electrochemical reductions indicating their good electrochemical sta-bility and electron transport in LECs The LEC ITOPEDOTPSS (40 nm)36[BMIM][PF6] (molar ratio 1035) (100 nm)Al emitted blue-green light centeredat 497 nm [32] with a maximum brightness of 39 cdm2 and a peak current effi-ciency of 84 cdA under 40 V White LEC were assembled based on theblue-green LEC which showed a peak current efficiency of 112 cdA at 35 V
He and co-authors further developed a 22rsquo-biimidazole N^N ligand and used itto assemble complex 37 [32] The biimidazole ligand exhibits a strong
212 L He
electron-donating property because it involves two nitrogen atoms of which thelong pair electrons are conjugated into the ligand plane Indeed compared tocomplex 1 complex 37 showed a significantly destabilized (ca 06 eV) LUMO Inacetonitrile solution complex 37 emitted blue-green light peaked at 496 nm withconsiderable ppy-centered 3p-p character Nevertheless complex 37 showed a lowluminescent efficiency owing to the twisting of the biimidazole plane caused bysteric hindrance between the two methyl groups The LEC ITOPEDOTPSS(40 nm)37 [BMIM][PF6] (molar ratio 1035) (100 nm)Al emitted green lightpeaked at 524 and 497 nm with a maximum brightness of 68 cdm2 and a peakcurrent efficiency of 04 cdA at 50 V [32] The low luminescent efficiency highLUMO level and irreversible electrochemical reduction of complex 37 should beresponsible for its low LEC performance
Sunesh and co-authors developed complexes 38 and 39 with apyridine-imidazole ancillary ligand of 2-(4-ethyl-2-pyridyl)-1H-imidazole [34] Inacetonitrile solution complexes 38 and 39 emitted blue-green and blue light
N
IrN
N
+
R=H 35 R=F 36
R
R
R
R
N NN
Ir
N
+
37
N N
NN
N
IrN
N
+
R=H 38 R=F 39
R
R
R
R
N NH
N
IrN
N
[X]-+
R=adamantyl X=[PF6] 40a R=adamantyl X=[BF4] 40b R=CH2Ph X=[PF6] 41aR=CH2Ph X=[BF4] 41b
F
F
F
F
NN N
RN
IrN
N
[X]-+
R=Ph X=[PF6] 42a R=Ph X=[BF4] 42b R=biphenyl X=[PF6] 43aR=biphenyl X=[BF4] 43b
F
F
F
F
NN N
RN
IrN
N
+
R1=Me R2=H 44 R1=R2=F 45
R2
R1
R1
R2
NN
NCN
N
IrN
N
+
R1=Me R2=H 46 R1=R2=F 47
R2
R1
R1
R2
NN
N3
4
[PF6]- [PF6]- [PF6]-
[PF6]-
[PF6]-
Fig 84 Ionic iridium(III) complexes using ancillary ligands beyond the bpy skeleton
8 Recent Advances on Blue-Emitting Iridium(III) Complexes hellip 213
respectively The emission showed dominant ligand-centered 3p minus p character incontrast to the dominant 3LLCT3MLCT character for the emission of complexes35ndash36 Therefore the N-H in 2-(4-ethyl-2-pyridyl)-1H-imidazole appears to have astronger electron-donating property than N-phenyl in pyim In neat films theemission of complexes 38ndash39 was red-shifted by ca 40 nm relative to that insolution The LEC ITOPEDOTPSS (40 nm)38 or 39 (75 nm)Al emitted greenand blue-green light with the emission maximums at 522 and 500 nm for com-plexes 38 and 39 respectively [34] Driven under sweeping voltages theblue-green LEC incorporating complex 39 gave a peak current efficiency of088 cdA
Mydlak and co-authors developed ancillary ligands with a skeleton of 2-(1H-123-triazol-4-yl)pyridine [35] Similar to the pyridine-pyrazole ligands thesepyridine-triazole ligands had high LUMO levels Using the pyridine-triazole typeN^N ligands and the dfppy C^N ligands complexes 40ndash43 were developed whichemitted blue light peaked around 452 and 484 nm in dichloromethane solution [35]Nevertheless complexes 40ndash43 showed irreversible or not fully reversible oxidationand reduction processes LECs were fabricated with a structure of ITOPEDOTPSS(100 nm)40ndash43 TBATf (molar ratio 11) (70 nm)Al with tetrabutylammoniumtrifluoromethanesulfonate (TBATf) doped as an ionic liquid The devices emittedblue light featuring a spectrum with a maximum and a shoulder at around 488 and460 nm respectively The EL spectra exhibited only small red-shifts with respect tothe PL spectra of the emissive layers Under 50 V the LECs showed maximumbrightness of 145 minus 449 cdm2 and half lifetimes of 34 minus 38 min
Chen and co-authors developed pyridine-triazole ancillary ligands with askeleton of 2-(4H-124-triazol-3-yl)pyridine and used them to construct blue-greenandor blue-emitting complexes 44ndash47 [36] In acetonitrile solution complexes 44and 46 using the ppy C^N ligands emitted blue-green light peaked at 480 nm andcomplexes 45 and 47 using the dfppy C^N ligands emitted blue light peaked at454 nm From the solution to the neat films the emission exhibited only smallred-shifts Reversible oxidation and quasi-reversible reduction were observed forthe complexes in solution The LECs ITOPEDOTPSS44ndash47 14 wt [BMIM][PF6] (100 nm)Al gave blue-green light for complexes 44 and 46 and sky bluelight for complexes 45 and 47 [36] Under 45 V LECs with complex 44 showed amaximum brightness of 308 cdm2 and a peak current efficiency of 68 cdA andthe values for the LEC with complex 45 were 202 cdm2 and 22 cdA It wasobserved that LECs with complex 44 showed faster response larger current densityhigher brightness and higher efficiency than LECs with complex 46 Similar resultswere also found between complexes 45 and 47 The enhanced performances for theLECs incorporating complexes 44 and 45 were tentatively ascribed to the higher ionmobility in the LECs which could be rendered by the cyanogen groups in thecomplexes
214 L He
823 Using Ancillary Ligands with StrongLigand Field Strength
For iridium(III) complexes the ligand field of the ligands affects directly the d-d splitting on the iridium ions thus impacting the d-involved HOMOs the energyof the 3MC states and the excited-state properties of the complexes Li andco-authors systematically investigated the influence of the ligand field strength ofthe ancillary ligands on the energy levels and emission properties of iridium(III)complexes [37] They found that compared to the ligands with weak ligand fieldstrength the ligands with strong ligand field strength split the d orbitals of theiridium ions to a larger extent stabilizing the HOMO levels and blue-shifting theemission of the complexes They also concluded that the increase of the emissionenergy led to a decrease on the non-radiative decay rate based on the energy gaplaw and a large d-d splitting destabilized the 3MC states away from the emittingtriplet state further suppressing the non-radiative deactivation for the emittingtriplet state Therefore ancillary ligands with strong ligand field strength canblue-shift the emission and meanwhile can enhance the luminescent efficiency ofthe complex On the other hand ancillary ligands with strong ligand field strengthreduce the MLCT character in the emitting triplet excited-state leading to decreasedradiative decay rates and thus long-lived triplet states [37]
The use of ancillary ligands with strong ligand field strength accounts for animportant avenue to blue-shift the emission of ionic iridium(III) complexes Theancillary ligands used for this purpose are in general monodentate ligands such aspyrazole CNminusCN-R phosphine and carbene For complexes including theseancillary ligands the HOMOs distribute on the phenyl rings of the ppy-type C^Nligands and the iridium ions similar to that in complex 1 however the LUMOs aremainly delocalized over the pyridine moieties of the ppy-type C^N ligands ratherthan on the ancillary ligands The migrations of the LUMOs from the ancillaryligands to the main ligands should be ascribed to the high energy levels of theunoccupied molecular orbitals on the ancillary ligands Therefore compared tocomplex 1 these complexes have intrinsically destabilized LUMO levels widenedenergy gaps and blue-shifted emission In coincidence with the LUMO migrationthe lowest emitting triplet states of the complexes usually exhibit mixedC^N-centered 3p minus p and 3MLCT (Ir C^N) character in contrast to the 3LLCT(ppy bpy)3MLCT (Ir bpy) character for complex 1 Hereafter all thecomplexes described in this section are displayed in Figs 85 and 86
The cyanide anion (CNminus) is a known monodenate ligand with strong ligandfield strength Nazeeruddin et al and Li et al used it for constructing anioniciridium(III) complexes [Ir(ppy)2(CN)2]
minusA+ with tetrabutylammonium as thecounter cation (A+) [37 38] Chen and co-authors developed an anionic iridium(III)complex 48 with a sodium ion as the counter cation [39] Compared to complex 1complex 48 showed a significantly destabilized (09 eV) LUMO level In ace-tonitrile solution it emitted blue-green light peaked at 472 and 502 nm with a high
8 Recent Advances on Blue-Emitting Iridium(III) Complexes hellip 215
luminescent efficiency of 070 and a relatively long excited-state lifetime (s) of40 ls The LEC ITOPEDOTPSS48 (100 nm)Al was non-emissive under elec-trical biases which was attributed to the poor film quality of the emissive layercaused by the poor solubility of complex 48 in organic solvent Dumura andco-authors developed an analogous complex 49 and fabricated a LEC ITOPEDOTPSS (40 nm)poly(N-vinylcarbazole) (PVK) (60 nm)49 (50 nm)Al (60 nm) [40]Because of the insertion of the PVK film the pinholes could be avoided and
N
IrCN
CNN
Na+
R=Me 48 R=H 49
N
IrN
N
A-+
FF
R=CH3 A=[PF6] 50a R=CH3 A=[BF4]50bR=C4H9 A=[PF6] 51a R=C4H9 A=[BF4] 51b
F
F
N
N
N
R
R
N
Ir N
NN
+
R1
R1
R1=R2=H 52 R1=F R2=H 53R1=F R2= COOCH3 54
R1
R1
R2
R2N NN
Ir
N
N
NN
+
FF
55
F
F
N
RR
NN
Ir
N
N
N
+
FF
56
F
F
N
N
N
[PF6]- [PF6]- [PF6]-
Fig 85 Ionic iridium(III) complexes using ancillary ligands with strong ligand field strength
IrN
[X]-+
N
N
NN
N
N
N
n-Bun-Bu n-Bu
n-Bu
X=I 57 X=[PF6] 58
N Ir
P P
N
[PF6]-
+
F
FF
F
P
P
O
PhPh
Ph PhP
P
O
PhPh
Ph PhP
PPhPh
Ph Ph
59 60 61P P =
xantphos dpephos dppe
Fig 86 Ionic iridium(III) complexes using ancillary ligands with strong ligand field strength
216 L He
meanwhile hole injection was facilitated for the LEC The LEC emitted green-bluelight centered at 488 nm with xy CIE color coordinates of 029045 and a peakcurrent efficiency of 0059 cdA under sweeping voltages
Yang and co-authors developed blue-emitting complexes 50ndash51 usingbis-carbene ancillary ligands and the dfppy C^N ligands [41] In dichloromethanesolution complexes 50ndash51 emitted blue light peaked at 452 and 482 nm arisingfrom the dfppy ligands and gave luminescent efficiencies of around 03 In neatfilms the complexes still emitted blue light exhibiting only small red-shifts withrespect to their PL spectra in solution but suffered from severe concentration-quenching The LECs ITOPEDOTPSS (100 nm)50 or 51 TBATf (molar ratio11) (70 nm)Al gave blue-green light peaked at 488 nm [41] The best xy CIEcolor coordinates for the EL were 020034 for complex 51a At 50 V the LECsgave peak current efficiencies at 037 minus 085 cdA and maximum brightness at134 minus 257 cdm2 One reason for the relatively low device performances could bethe less efficient electron transport as indicated by the irreversible electrochemicalreductions of complexes 50ndash51
By using carbene-pyridine type ancillary ligands Kessler and co-authorsdeveloped blue-green to blue-emitting complexes 52ndash55 [42] Due to the very highLUMO levels of the carbene-pyridine ligands the emission of the complexesoriginated from the C^N ligands Upon adding more electron-withdrawing groupsto the phenyl rings of the C^N ligands the emission was more blue-shifted with theemission peak blue-shifted from 471 to 450 to 447 and to 435 nm for complexes52 53 54 and 55 respectively In the solution complex 55 exhibited the highestefficiency of 020 and the longest s of 85 ls For complexes 52ndash55 the 3MC statesconcerning the rupture of the Ir-N bond between the iridium and the carbene-pyridine ligand accounted for an efficient non-radiative deactivation channel for theemitting triplet states The LEC ITOPEDOTPSS (100 nm)53 [BMIM][PF6](molar ratio1033) (150 nm)Al gave green emission centered at 512 nm [42] Thered-shift between PL and EL was tentatively attributed to either a polarization effectcaused by electrical excitation or a change of the emissive excited-state insolid-state
By using a bis-carbene ancillary ligand and 2rsquo6rsquo-difluoro-23rsquo-bipyridine as theC^N ligand Meier and co-authors developed a deep-blue-emitting complex 56[43] In acetonitrile solution complex 56 emitted blue light with emission peaksaround 440 and 480 nm and a luminescent efficiency of 013 In thin films complex56 still emitted blue light with only small red-shifts relative to those recorded insolution but showed severe concentration-quenching Notably the bis-carbeneancillary ligands eliminated the low-lying 3MC states existing in complexes 52-55with the pyridine-carbene ancillary ligands The LEC ITOPEDOTPSS (100 nm)56[BMIM][PF6] (molar ratio1033) (100 nm)Al emitted greenish light centeredat 500 nm [43] The EL was gradually red-shifted under the continuous electricalexcitation as revealed by the change of xy CIE color coordinates from 027043 to036045 after ten minutes of operation due to the broadening of the EL spectrumUnder 60 V the LEC reached a maximum brightness of 15 cdm2 and a peakcurrent efficiency of 176 cdA Biased at a constant current of 25 mAcm2 the
8 Recent Advances on Blue-Emitting Iridium(III) Complexes hellip 217
LEC showed a better performance with a maximum brightness of 113 cdm2 and apeak current efficiency of 47 cdA The LECs showed very limited stability withhalf lifetimes of only several minutes
For complexes 50ndash56 with carbene-type ancillary ligands the non-carbene C^Nligands dominated the emission because of their smaller energy gaps compared tothe carbene ancillary ligands Darmawan and co-authors followed a differentapproach employing a tridentate bis-carbene ligand 13-phenylene-jC2)bis(1-butylimidazol-2-ylidene) to construct homoleptic cationic bis-pincer Ir(III)complexes 57ndash58 [44] The tridentate bis-carbene ligands had very strong ligandfield strength and formed a rigid bis-terdentate coordination with the central iridiumion For such complexes the emission was mainly dominated by the tridentatebis-carbene ligands In acetonitrile solution complexes 57 and 58 gave similarnear-UV emission peaked around 384 and 406 nm with luminescent efficiencies of041 038 and long s of 89 94 ls respectively In films both complexes 57 and58 showed a new low-energy emission at around 500 nm along with the blueemission at around 400 nm This was tentatively attributed to the formation of newtrapping species associated with the counter anions The LEC ITOPEDOTPSS(80 nm)PMMA 50 wt 58 (80 nm)Al emitted near-UV light peaked at 386 and406 nm [44] The EQE of the LEC was below 1 This LEC represents the bluestLEC incorporating iridium(III) complexes reported so far
With strong r-donating character phosphine has high ligand field strength [1445] Recently Martir and co-authors conducted a systematical investigation on thestructure-property relationship for cationic iridium(III) complexes with differentbidendtate phosphine ancillary ligands [45] Complexes 59-61 used biphosphineligands of xantphos dpephos and dppe as the ancillary ligands and 2-(46-difluorophenyl)-4-mesitylpyridine as the C^N ligand (Fig 86) In acetonitrilesolution complexes 59 60 and 61 emitted blue-green light peaked at 489 484 and471 nm respectively Complex 61 showed the bluest emission the highest lumi-nescent efficiency (052) and the longest s value (135 ls) because thefive-membered rigid chelation ring in complex 61 rendered stronger Ir-P bonds (thusstronger ligand field strength) than the flexible eight-membered chelation ringsformed in complexes 59 and 60 In neat films the emission exhibited only smallred-shifts compared to the PL in solution and moderate concentration-quenching ofluminescence due to the bulky mesityl and biphosphine ligands The LECITOPEDOTPSS (80 nm)59ndash61 [BMIM][PF6] (molar ratio 1025) (100 nm)Alhowever showed very low luminescence level even under a high-pulsed current of765 A mminus2 which was attributed to the irreversibility of the phosphine ligandoxidation and the high driving current density [45] The host-guest LECITOPEDOTPSS (80 nm)host complex 61 (mass ratio 04504501) (100 nm)Al showed sky blue emission peaked at 479 nm with a maximum brightness around8 cdm2 under a pulsed current density of 100 A mminus2 [45]
218 L He
824 Using Cyclometalating Ligands Beyond the ppySkeleton
The C^N ligands play a critical role on the emission of iridium(III) complexesbecause the HOMOs distribute partially on their phenyl rings The typical C^N ligandused for iridium(III) complexes is ppy The complexes with the ppy C^N ligandssuch as complex 1 have narrow energy gaps Attaching electron-withdrawinggroups such as fluorine to the phenyl rings of ppy blue-shifts the emission but itdeteriorates the stability of the complexes in LECs (vide supra) Other C^N ligandsbeyond the ppy skeleton are thus desired for blue-emitting complexes Hereafter allthe complexes described in this section are displayed in Figs 87 88 and 89
Researchers have replaced the pyridine ring in the ppy with a nitrogen-richfive-membered heterocycle such as pyrazole triazole or tetrazole to develop newionic iridium(III) complexes In most cases such modifications stabilize the HOMOto some extent thus blue-shifting the emission of the complex [33] The HOMOstabilization is usually caused by the high ligand field strength or theelectron-deficient characteristic of the nitrogen-rich five-membered heterocycleTamayo and co-authors developed complexes 62ndash63 by using the phenyl-pyrazoleC^N ligands and the dtb-bpy N^N ligand [46] With the 1-phenyl-1H-pyrazole(ppz) C^N ligand complex 62 emitted yellow light centered at 555 nm in ace-tonitrile solution blue-shifted by about 30 nm relative to that of complex 2 owingto the HOMO stabilization (ca 01 eV) caused by ppz With the fluorinated ppzie 1-(24-difluorophenyl)-1H-pyrazole (dfppz) as the C^N ligand complex 63
N
IrN
N
N[PF6]-+
R=H 62 R=F 63
N
N
R
R
R
RN
IrN
N
N
+
RR=22-biphenyl 64RR=4-CH3OPh 4-CH3Oph 65RR=C2H5 C2H5 66
N
N
F
F
F
F
R
R
N
IrN
N
N
+
F
F
N
N
R2
R1=R2=H 67 R1=H R2=ph 68R1=R2=ph 69
F
F
R1
N
IrN
N
N
+
F
F
R1
R1N
N
R3
R1=R2=H R3=phenyl 70R1=tBu R2=H R3=phenyl 71R1=tBu R2=R3=phenyl 72R1=N(CH3)2 R2=R3=H 73
F
F
R2
N
IrN
N
+
N
N
F
F
F
F
N N
74
[PF6]- [PF6]-
[PF6]-[PF6]-
Fig 87 Ionic iridium(III) complexes using cyclometalating ligands beyond the ppy skeleton
8 Recent Advances on Blue-Emitting Iridium(III) Complexes hellip 219
emitted blue-green light peaked at 495 nm in acetonitrile solution blue-shifted by17 nm compared to that of complex 6 The LEC ITOPEDOTPSS (40 nm)63(70 minus 90 nm)Al gave blue-green light centered at 492 nm with xy CIE colorcoordinates of 020040 and a peak EQE of 46 under 30 V [46]
Since reported complex 63 has been widely employed as a blue-green emitterfor LECs Liao and co-authors doped a hole-trapper of 33rsquo-diethyl-22rsquo-oxathia-carbocyanine iodide (DOTCI) in the blue-green LECs incorporating complex 63[47] The LEC structure was ITOPEDOTPSS (30 nm)63 DOTCI [BMIM][PF6][mass ratio (80-x) x 20] (200 nm)Ag Compared to complex 63 DOTCl had amuch higher HOMO level but a similar LUMO level thus serving as a hole-trapperin the emissive layer Because complex 63 preferred to transport holes dopingDOTCl impeded the hole mobility and improved the balance between the hole andelectron transport which moved the carrier recombination zone toward the center ofthe emissive layer and improved the device efficiency Upon doping 001 wtDOTCl the maximum brightness and the peak EQE at 33 V were increased from109 cdm2 to 166 cdm2 and from 906 to 1275 respectively Using complex63 the same group also assembled a white LEC of glassCCLITO (120 nm)PEDOTPSS (30 nm)63 20 wt [BMIM][PF6] (350 nm)Ag (100 nm) byinserting a red color-conversion layer (CCL) between the ITO and the glass sub-strate [48] which showed a maximum brightness of 32 cdm2 and a peak EQE of125 under 37 V
As shown below dfppz has been widely used for constructing blue-emittingiridium(III) complexes Su and co-authors developed dfppz-based iridium(III)emitters 64-66 with 45-diazafluorene (daf) as the N^N ligand [49] In dichlor-omethane solution complexes 64-66 emitted blue-green light centered at 491ndash499 nm with high luminescent efficiencies of 046ndash066 In neat films the lumi-nescent efficiencies maintained at 020ndash028 These high luminescent efficienciesshould be ascribed to the rigidity of the daf ligands Complex 66 showed the bluestemission among the three complexes and was chosen to assemble a white LECwhich showed a maximum brightness of 25 cdm2 and a peak EQE of 40 under29 V [49]
Chen and co-authors developed another series of dfppz-based blue-green emit-ters 67ndash69 with 45-diaza-99rsquo-spirobifluorene as the N^N ligand in which one ortwo pendant phenyl rings were attached ortho- to the Ir-N bonds [50] In ace-tonitrile solution complexes 67ndash69 emitted blue-green light centered at 500ndash505 nm with luminescent efficiencies of 038ndash046 In neat films the emission wasred-shifted to 510ndash512 nm The intermolecular p-p stacking interactions wereformed between the pendant phenyl rings of the ancillary ligands and the phenylrings of dfppz as observed from the crystal structures [5] The LECs ITOPEDOTPSS67ndash69 [BMIM][PF6] (mass ratio 8020) (400 nm)Al showed green-blueemission with xy CIE color coordinates of 023047 028050 and 028054 forcomplexes 67 68 and 69 respectively [50] The brightness reached the maxima of254 58 and 106 cdm2 with peak EQEs of 082 113 and 143 and halflifetimes of 141 585 and 102 min for devices incorporating complexes 67 68 and69 at 34 V respectively The LECs with complexes 68 and 69 showed slower
220 L He
response and lower current densities presumably due to the suppressed carrierinjectiontransport in the supermolecular-caged structures of densely extendedp-stacking and low ionic mobility Although containing intramolecular p-p stackinginteractions complexes 67ndash69 showed limited stability in LECs
Baranoff and co-authors developed dfppz-based blue-green emitters 70ndash73 usingbpy-type ancillary ligands in which pendant phenyl rings were attached ortho- tothe Ir-N bonds (complexes 70ndash72) or dimethylamino substituents were attachedpara to the Ir-N bonds (complex 73) [51] For complexes 70ndash72 strongintramolecular p minus p stacking interactions were formed In acetonitrile solutioncomplexes 70ndash73 emitted blue-green light with the emission maximums at 517505 501 and 493 nm respectively The complexes showed very high luminescentefficiencies of 06 minus 10 in diluted films The LECs ITOPEDOTPSS (90 nm)70-73[BMIM][PF6] (molar ratio = 1025) (90 nm)Al emitted yellow or orangelight peaked at 545ndash574 nm [51] Because the PL spectra of the emissive layerswere centered around 500 nm the large red-shifts between the EL and PL should becaused by electrical excitation The underlying reasons for such large red-shiftswere unclear and could be related to the morphological effects andorlight-out-coupling effects
He and co-authors developed a dfppz-based blue-green emitting complex 74 byusing a bulky pyridine-imidazole ancillary ligand [52] In acetonitrile solutioncomplex 74 emitted blue-green light peaked at 494 and 472 nm This emission wasblue-shifted compared to the emission (497 nm) from complex 36 with the dfppyC^N ligand In neat films the PL of complex 74 was only slightly red-shifted andthe luminescent efficiency maintained as high as 054 due to the significantlysuppressed intermolecular interactions in the film caused by the bulky 4-tritylphenylgroup attached at the ancillary ligand The LEC ITOPEDOTPSS (40 nm)74[BMIM][PF6] (molar ratio 11) (100 nm)Al featured blue-green light with xyCIE color coordinates of 022041 [52] Under 32 V the LEC reached a maximumbrightness of 145 cdm2 and a peak EQE of 76 The efficiency was largelyenhanced compared to the LEC incorporating complex 36 due to the suppressedluminescence concentration-quenching for complex 74 White LECs assembledwith complex 74 showed a peak current efficiency of 114 cdA and a peak powerconversion efficiency of 112 lmW [52]
Shan and co-authors developed a dfppz-based blue-green emitting complex 75using a pyridine-triazole ancillary ligand [53] Using complex 75 Wu andco-authors assembled a flexible LEC of PETITO75 [BMIM][PF6] (mass ratio21)Al where PET was polyethylene terephthalate as a flexible substrate [54]The LEC gave blue-green light centered at 503 nm with xy CIE color coordinatesof 025048 Under 50 V the blue-green LEC reached a maximum brightness of365 cdm2 and a peak current efficiency of 107 cdA The flexible white LEC werefabricated with complex 75 showing a peak current efficiency of 98 cdA at 70 VThe efficiencies of both the blue-green and white LECs featured no obviousdegradation after bending the devices at 10 mm curvature radius for 200 times
8 Recent Advances on Blue-Emitting Iridium(III) Complexes hellip 221
Other electron-withdrawing groups rather than fluorine such as sulfonyl groups(vide supra) have been attached to ppz for fluorine-free blue-emitting complexesTordera and co-authors developed complexes 76ndash81 by using methylsulfonyl sub-stituted ppz as the C^N ligands and bpy derivatives as the N^N ligands [18] Inacetonitrile solution complexes 76ndash81 emitted blue-green to green light centeredbetween 492ndash518 nm with luminescent efficiencies of 038ndash080 Complex 79exhibited the bluest emission centered at 492 nm due to the attachment of theelectron-donating dimethylamino group onto bpy LECs were fabricated incorpo-rating complexes 76ndash78 and 80ndash81 with a configuration of ITOPEDOTPSS(90 nm)complex[BMIM][PF6] (molar ratio 1 025) (90 nm)Al [18] The LECsfeatured green light peaked around 500 nm for complexes 77 and 80 and 515 nm forcomplexes 76 78 and 81 Upon using a driven mode based on applying pulsedcurrent densities of 25-100 Am2 LECs based on complexes 76 and 77 gave betterperformances with peak current efficiencies of around 15 cdA Although complexes76ndash81 were fluorine-free and contained intramolecular p-p stacking interactions theLECs showed limited stability The longest half lifetime of 54 min at the maximumbrightness of 334 cdm2 was achieved by the LEC incorporating complex 76
Similar to the phenyl-pyrazole ligand the phenyl-triazole ligand contains anitrogen-rich five-membered heterocycle and has been employed as the C^N ligandfor ionic iridium(III) complexes For example ligands with scaffolds of1-decyl-4-phenyl-1H-123-triazole [55] 2-phenyl-2H-123-triazole [56] or1-benzyl-4-phenyl-1H-123-triazole (phtl) [57] have been tested for this purposeThe phtl-type ligands have been shown to blue-shift the emission of the complex by
N
IrN
N[PF6]-+
N
N
F
F
F
F
NNN
75
N
IrN
N
N
+
MeO2S
MeO2S
R
RN
N
R=H 76 R=tBu 77R=SCH3 78 R=N(CH3)2 79
N
IrN
N
N+
MeO2S
MeO2S
R1
R2N
N
Ph
R1=R2=tBu 80 R1=H R2=SCH3 81
IrN
+
NN
N
NN
N
R1=F R2=H R3=Ph 82R1=F R2=H R3=CH2Ph 83R1=R2=F R3=Ph 84R1=R2=F R3=CH2Ph 85
CH2Ph
CH2Ph
NN N
R3
R2
R1
R1
R2
IrN
N
+
R
RN N
NN
NN
NN
R=H 86 R=tBu 87
[PF6]-
[PF6]- [PF6]- [BF4]-
Fig 88 Ionic iridium(III) complexes using cyclometalating ligands beyond the ppy skeleton
222 L He
10ndash20 nm as compared to ppy [55 57] However the extent of the blue-shift wassmaller than that caused by ppz With fluorinated phtl as the C^N ligand andderivatives of pyridine-123-triazole (pytl) as the N^N ligands Fernaacutendez-Hernaacutendez and co-authors developed blue-emitting complexes 82ndash85 [58] Inacetonitrile solution complexes 82ndash85 emitted sky blue or deep-blue light peakedat 487 485 461 and 452 nm respectively The excited-states featured a dominant3LLCT (phtl pytl)3MLCT (Ir pytl) character In neat films the emission wasstill in the blue region centered at 451ndash473 nm Nevertheless the blue emissionfrom complexes 82ndash85 was very weak in both solutions or diluted films This wastentatively attributed to the luminescence-quenching caused by the non-radiative3MC states The LECs ITOPEDOTPSS (90 minus 100 nm)82ndash85LiF (1 nm)Alemitted yellowish green to blue light which was largely red-shifted compared to thePL in solution or films [58] This red-shift between the EL and PL was tentativelyascribed to the excimer formation in the film under electrical excitation The LECbased on complex 83 gave the bluest emission centered at 487 nm with xy CIEcoordinates of 026036 Under 90 V the LECs showed maximum brightness at3 minus 25 cdm2 and peak current efficiencies at 03 minus 14 cdA
The phenyl-tetrazole ligand contains a heterocycle with four nitrogen atoms andhas also been employed as the C^N ligand for ionic iridium(III) complexes Montiand co-authors developed complexes 86-87 by using the phenyl-tetrazole C^Nligand and the bpy-type N^N ligands [59] With a high ligand field strength thephenyl-tetrazole ligands stabilized the HOMOs of the complexes by 03 eV ascompared to the ppy ligands which was larger than that caused by ppz (01 eV) Inacetonitrile solution complexes 86 and 87 emitted green light peaked at 545 and530 nm respectively They were significantly blue-shifted (45-60 nm) compared tothat of complex 1 This blue-shifting effect was also larger than that caused by ppzThe LEC ITOPEDOTPSS (80 nm)87 [BMIM][PF6] (90 nm)Al emittedyellow-green light centered at 552 nm [59] which was red-shifted compared to thePL in solution owing to strong intermolecular interactions in the emissive layerAlthough no fluorine substitutions were presented in complex 87 the LEC showeda limited stability largely due to the irreversible oxidation of the complex inducedby the tetrazole heterocycle
The phenyl-imidazole ligand contains a nitrogen-rich imidazole ring He andco-authors examined the use of phenyl-imidazole as the C^N ligand for cationiciridium(III) complexes [60] Different from the phenyl-pyrazole or phenyl-triazolethe ligand with a scaffold of 2-phenyl-1H-imidazole did not stabilize the HOMOs ofthe complexes Instead the phenyl-imidazole ligands destabilized simultaneouslythe HOMO and LUMO levels presumably due to the electron-donating property ofthe imidazole ring within the complexes With the 12-diphenyl-1H-imidazole(dphim) C^N ligand and the bpy N^N ligand complex 88 gave orange-red light verysimilar to that of complex 1 Therefore the nitrogen-rich five-membered heterocycledoes not always blue-shift the emission of the complex and the blue-shifting effect isdetermined by the relative positions of the nitrogen atoms in the heterocycle [33]
8 Recent Advances on Blue-Emitting Iridium(III) Complexes hellip 223
With the phenyl-imidazole C^N ligand and the pzpy N^N ligand complexes 89 and90 were developed which emitted blue-green light centered around 505 nm inacetonitrile solution with dominant 3LLCT (C^N N^N)3MLCT (Ir N^N))character The LECs ITOPEDOTPSS (40 nm)89 or 90 [BMIM][PF6] (molarratio 1 05) (100 nm)Al emitted green-blue light centered around 510 nm withmaximum brightness of 402 178 cdm2 and peak current efficiencies of 6324 cdA for complexes 89 and 90 respectively under 40 V
Other heterocycles such as oxadiazole (oxd) have been used to replace thepyridine ring in ppy Wang and co-authors developed complexes 91ndash92 by using25-diphenyl-134-oxadiazole (dphoxd) as the C^N ligand [61] Compared to thepyridine ring in ppy the oxd heterocycle was strongly electron-deficient whichstabilized the HOMO levels of the complexes by more than 04 eV In acetonitrilesolution complexes 91 and 92 emitted similar blue-green light peaked around 490and 520 nm with the same character of mixed dphoxd-centered 3p minus p and3MLCT (Ir dphoxd) irrespective of the different ancillary ligands (bpy or pzpy)used for the complexes The emission was blue-shifted by about 100 nm relative tothat of complex 1 owing to the HOMO stabilization induced by oxd The LECsITOPEDOTPSS (40 nm)91 or 92 [BMIM][PF6] (molar ratio 1 08) (100 nm)Al showed yellow light centered at 556 nm for complex 91 and green-blue lightpeaked at 492 530 nm for complex 92 [61] The large red-shift between the EL andthe solution PL for complex 91 was attributed to the change of the emitting tripletstate from dphoxd-centered 3p minus p in solution to the charge-transfer state in theconcentrated film similar to that occurred in complex 5 [13] The green-blue LEC
Ir
N N
[PF6]-
+
O
NN N
N
O N N = bpy 91
NIr
N
NBu3P
PBu3
PBu3
PF6-
PF6- PF6-
93
IrN
N
[PF6]-
N
N
N
N
N
R
RR=Ph 89 R=n-Bu 90
+
IrN
[PF6]-
N
N
N
NPh
Ph88
+
N
N N = pzpy 92
Fig 89 Ionic iridium(III) complexes (except 93) using cyclometalating ligands beyond the ppyskeleton
224 L He
incorporating complex 92 showed a maximum brightness of 30 cdm2 and a peakcurrent efficiency of 52 cdA under 35 V
Complexes 1ndash92 all contain two anionic C^N ligands and their ionic nature isrendered by the ancillary ligands Bolink and co-authors developed an ionic iridium(III) complex 93 by ionizing the neutral complex Ir(ppy)3 through attaching cationictri-n-butylphosphonium side groups in the periphery of the ppy ligands [62] Inacetonitrile solution complex 93 emitted blue-green light peaked at 480 nm whichwas blue-shifted by 35 nm relative to that of Ir(ppy)3 owing to the HOMO sta-bilization caused by the electron-withdrawing tri-butyl phosphor group The LECITO9320 wt PMMA (100ndash200 nm)Au emitted green-blue light with a maxi-mum at 487 nm right after the turn-on of the device [62] However the EL wasgradually red-shifted to the yellow region upon the continuous operation of theLEC This red-shift of EL was tentatively attributed to the chemical degradation ofthe complex under the electrical excitation
83 Conclusion and Outlook
831 Current Status
So far a wide variety of blue-emitting iridium(III) complexes have been developedand applied for LECs The performances of the blue or blue-green LECs with theEL maximums around or below 500 nm have been summarized in Table 81
As summarized in Table 81 great progress has been made to improve the colorpurity efficiency and brightness turn-on time and stability of blue LECsNevertheless problems still remain and further improvements to these parametersare needed for applications We provide a brief description as follows
1 Color purity
Tremendous efforts have been devoted to blue-shifting the emission of ionic iridium(III) complexes for LECs So far good color purity has been achieved Forexample Darmawan et al reported the near-UV emitting complexes (57ndash58) andLECs which showed EL maximums at 386 and 406 nm [44] In general forconstructing warm-white LECs blue-green or sky blue LECs would fulfill therequirements of which the xy CIE color coordinates should be around or betterthan 020040 [32 48 49 52 54] Many blue or blue-green LECs have alreadyreached this standard (Table 81)
Nevertheless blue-emitting complexes do not always lead to blue LECs Manycomplexes are good blue or blue-green emitters in solution or in diluted films buttheir corresponding LECs feature electroluminescence responses that are red-shiftedto the green or yellow region [12 13 20 21 23 26 29 30 42 51 58 59 61]This red-shift between the PL and EL can be caused by several factors such as(i) the downward shift of the triplet states as a result of the strong intermolecular
8 Recent Advances on Blue-Emitting Iridium(III) Complexes hellip 225
Tab
le81
Summaryof
theblue
orblue-green
LECsincorporatingiridium(III)complexes
LECa
Driving
mod
et on(m
in)
Bmax
(cdm
2 )Eff m
ax(cdA)
EQEm(
)t 12(m
in)
ELpeak
(nm)
CIE
(xy)
Ref
ITO4
(70ndash80
nm)
Au
minus60V
nr33
nr016
3050
0(020051)
[12]
ITO20(100
nm)
Ag
35V
nr20
0043
020
nr52
049
1nr
[24]
ITOPEDOTPSS
22
(75nm
)Al
50V
432
52096
036
300b
48650
8(027050)
[26]
ITOPEDOTPSS
23
(75nm
)Al
65V
426
39065
028
420b
46048
6(020028)
[26]
ITO22
IL(1035)
(120
nm)Al
50V
2494
43
16
nr48
651
2(029050)
[26]
ITOPEDOTPSS
24
(75nm
)Al
Sweeping
voltage
nrnr
046
nrnr
48851
6(028050)
[27]
ITOPEDOTPSS
25
(75nm
)Al
Sweeping
voltage
at05Vs
nrnr
038
nrnr
46349
1(024042)
[27]
ITOPEDOTPSS
28
Al
Sweeping
voltage
at05Vs
nrnr
031
nrnr
503
(021051)
[29]
ITOPEDOTPSS
29
Al
Sweeping
voltage
at05Vs
nrnr
034
nrnr
511
(024053)
[29]
ITOPEDOTPSS
30
ndash33IL
(1025)
(100
nm)Al
Pulsed
100
Am
205ndash13
49ndash14
1nr
02ndash04
24ndash62
492ndash49
653
0ndash54
7nr
[30]
ITOPEDOTPSS
34
IL
(105)
(100
nm)Al
30V
300
3787
34
950
48651
2(026048)
[31]
(con
tinued)
226 L He
Tab
le81
(con
tinued)
LECa
Driving
mod
et on(m
in)
Bmax
(cdm
2 )Eff m
ax(cdA)
EQEm(
)t 12(m
in)
ELpeak
(nm)
CIE
(xy)
Ref
ITOPEDOTPSS
36
IL(1035)
(100
nm)Al
40V
222
3984
34
600d
497
(025046)
[32]
ITOPEDOTPSS
39
(75nm
)Al
Sweeping
voltage
at05Vs
nrnr
088
nrnr
500
(024044)
[34]
ITOPEDOTPSS
40
ndash43TBATf(11)
(70nm
)Al
50V
20ndash14
315ndash45
nrnr
34ndash38
48846
0nr
[35]
ITOPEDOTPSS
44
14
wt
IL(100
nm)Al
40V
nr97
35
14
nr48
451
7(031053)
[36]
ITOPEDOTPSS
45
14
wt
IL(100
nm)Al
40V
nr53
26
27
nr45
848
8(021033)
[36]
ITOPEDOTPSS
46
14
wt
IL(100
nm)Al
45V
nr17
517
067
nr48
351
5nr
[36]
ITOPEDOTPSS
47
14
wt
IL(100
nm)Al
45V
nr45
026
027
nr45
748
6nr
[36]
ITOPEDOTPSS
50
aor
50b
TBATf
(11)(70nm
)Al
50V
45ndash10
621
6ndash25
7073ndash050
nr80ndash53
848
8(027043)
[41]
ITOPEDOTPSS
51
aor
51b
TBATf
(11)(70nm
)Al
50V
53ndash78
240ndash13
4037ndash085
nr98ndash16
748
845
6(020ndash022
034ndash038)
[41]
(con
tinued)
8 Recent Advances on Blue-Emitting Iridium(III) Complexes hellip 227
Tab
le81
(con
tinued)
LECa
Driving
mod
et on(m
in)
Bmax
(cdm
2 )Eff m
ax(cdA)
EQEm(
)t 12(m
in)
ELpeak
(nm)
CIE
(xy)
Ref
ITOPEDOTPSS
56
IL
(1033
)(100
nm)Al
60V
358
15176
nr30
500c
(028043)c
[43]
ITOPEDOTPSS
56
IL
(1033
)(100
nm)Al
25mAcm
2028
113
471
nr10
500c
(024040)c
[43]
ITOPEDOT
PSSPM
MA50
wt
58
(80nm
)Al
Sweeping
voltage
nrnr
nrlt1
0nr
40638
6nr
[44]
ITOPEDOT
PSSHost10
wt
61(100
nm)Al
Pulsed
100
Am
210
b78b
008
bnr
30b
479
nr[45]
ITOPEDOTPSS
63
(70ndash90
nm)Al
30V
100b
1700
105
46
nr49
2(020041)
[46]
ITOPEDOTPSS
63
20
wt
IL(200
nm)Ag
33V
4910
920
91
4049
0dnr
[47]
ITOPEDOTPSS
63
001
wt
DOTCI20
wt
IL(200
nm)Ag
33V
6216
630
128
4549
0dnr
[47]
ITOPEDOTPSS
67
ndash6920
wt
IL(400
nm)Al
34V
54ndash39
058ndash25
421ndash43
08ndash14
102ndash58
5nr
(023
028047
054)
[50]
ITOPEDOTPSS
74
IL
(11)
(100
nm)Al
32V
2914
518
376
50b
49447
4(022041)
[52]
PETITO75
33wt
ILAl
50V
2336
510
7nr
120
503
(025048)
[54]
(con
tinued)
228 L He
Tab
le81
(con
tinued)
LECa
Driving
mod
et on(m
in)
Bmax
(cdm
2 )Eff m
ax(cdA)
EQEm(
)t 12(m
in)
ELpeak
(nm)
CIE
(xy)
Ref
ITOPEDOTPSS
77
IL
(1025
)(90nm
)Al
50Am
2nr
890
155
49
30
490
520d
(027050)
[18]
ITOPEDOTPSS
80
IL
(1025
)(90nm
)Al
25Am
2nr
238
94
29
1249
053
0d(029050)
[18]
ITOPEDOTPSS
83
LiF
(1nm
)Al
90V
60
1303
014
4848
7(026036)
[58]
ITOPEDOTPSS
85
LiF
(1nm
)Al
90V
120
304
039
360
508
(028045)
[58]
ITOPEDOTPSS
89
IL
(105)
(100
nm)Al
40V
19
402
63
nr60b
508
(028048)
[60]
ITO93
20wt
PMMA
(100ndash
200nm
)Au
40V
083
32025
nr042
487
540c
(03405c
[62]
a Unlessotherw
isestatedtheionicliq
uid(IL)is
[BMIM
][PF
6]andtheratio
intheparenthesisis
themolar
ratio
nr
means
norepo
rted
datab Estim
ated
from
the
brightness-versus-tim
ecurvescThe
emission
was
red-shiftedup
onthecontinuo
usop
eration
d Estim
ated
from
theELspectraNoteThe
term
st ont 12B
maxE
ff maxrefer
toturn-ontim
elifetim
emaxim
umluminancemaxim
umefficiencyrespectiv
ely
The
definitio
nof
theseterm
sareprov
ided
inChap
1
8 Recent Advances on Blue-Emitting Iridium(III) Complexes hellip 229
interactions in the concentrated emissive layers (ii) the polarization of themolecular orbitals (iii) the excimer formation andor (iv) chemical changes of thecomplexes under the electrical excitation
Indeed for some blue or blue-green LEC their EL exhibited limited stabilityie the EL was gradually red-shifted upon the continuous operation of the LECs[43 62] The mechanism for such red-shift is not completely understood today Thecurrent ideas involve (i) the chemical degradation of the complexes under theelectrical excitation andor (ii) the intrinsic degradation of the devices
2 Efficiency and brightness
The efficiencies of the blue-green or blue LECs have been significantly improved(Table 81) For instance Liao et al reported a peak EQE of 128 for theblue-green LEC incorporating complex 63 [47] In general the EQEs of thin-filmlight-emitting devices are determined by EQE = vUηrn where v is the fraction ofexcitons that can decay radiatively (v = 1 for phosphorescent devices) U is thephotoluminescence quantum efficiency of the active layer ηr is the electron-holerecombination efficiency and n is the light-out-coupling efficiency of the device(n 02 for sandwiched thin-film devices)
To enhance the EL efficiency a high U value is required for the emissive layercomposed of the concentrated complexes To this end the complexes themselvesshould be highly emissive which requires a judicious tailoring on the molecularstructures so that high luminescent efficiencies are ensured while blue-shifting theemission In particular the luminescence-quenching 3MC states need to be desta-bilized away from the emissive triplet states [42] It is equally important to suppressthe concentration-quenching of the long-lived triplet states in the emissive layerswhich can be achieved by several approaches such as (i) increasing the sterichindrance of the complexes via using bulky ligands [45 50 52] (ii) physicallydiluting the complexes in the emissive layers via doping ionic liquid andor(iii) adopting a host-guest configuration [9 63]
To enhance the EL efficiency a balanced electronhole injection transport andrecombination is required [47] This task is more difficult than for green or red LECsbecause the blue or blue-green LECs have larger carrier injection barriers associatedwith the wide energy gaps of the complexes at the electrodeemissive layer inter-faces Ionic liquid are usually doped in blue LECs to reduce the large carrierinjection barriers for more balanced carrier injection leading to improved efficien-cies However this significantly reduces the device stabilitymdashsee Chaps 1 4 and 5for more details In general complexes with reversible electrochemical oxidationsand reductions showed better performances in LECs compared to those with irre-versible oxidations and reductions largely due to the smoother transports of holes(oxidation) and electrons (reduction) in LECs incorporating the former complexes
It has always been observed that the efficiencies of the blue or blue-green LECsdecrease rapidly upon the continuous operation of LECs This efficiency roll-offwith time should be related to the overall stability of the LECs ie the degradation
230 L He
of the devices under continuous electrical excitations which can be caused by(i) the chemical degradation of the complexes (ii) the migration of the carrierrecombination zone toward the electrodes and (iii) over-doping of the emissivelayersmdashsee Chaps 1 2 and 3 for more details
As shown in Table 81 most of the blue or blue-green LECs showed maximumbrightness below 100 cdm2 under operating voltages of 3minus4 V This brightness ismuch lower than that (hundreds of cdm2) of the orange-red LECs under similardriving voltages [1 9] The lower brightness for the blue or blue-green LECsshould be partially attributed to the more difficult carrier injection at theelectrodeemissive layer interfaces associated with the enlarged energy gaps of thecomplexes Increasing the operating voltages or driving the LECs with high-pulsedcurrent enhance the brightness at the expense of reduced device lifetimes
3 The turn-on time
The turn-on time of the LECs is defined as the time for the brightness to increase toa certain value upon application of the electrical excitationmdashsee Chap 1 for moredetails Here it is defined as the time for the brightness to increase to the peak valueAs shown in Table 81 the turn-on times for the blue or blue-green LECs rangefrom several seconds to several hours depending on the device structure themagnitude of the driving voltages or currents and the driving modes (constantvoltages versus pulsed currents) Doping ionic liquids provides additional mobileions for the electrochemical doping and increasing the driving force accelerates theion migration These two techniques shorten the response time but meanwhilereduce the stability of the LECs
4 Stability
The stability of LECs can be evaluated by their lifetimes which are defined as thetime for the brightness to decreases from the maximum to the half under thecontinuous electrical excitation As shown in Table 81 the blue or blue-greenLECs showed very limited lifetimes ranging from several minutes to several hoursBecause the peak brightness of the LECs is relatively low even worse stability isexpected for the blue or blue-green LECs as compared to the orange-red LECs
For LECs incorporating transition metal complexes the active complexes couldbe chemically decomposed after being attacked by foreign species such as waterunder the electrical excitation leading to limited lifetimes for the devicesmdashseeChap 7 for more details Following the success on enhancing the lifetime oforange-red LECs [5] several groups introduced intramolecular p-p stackinginteractions or hydrophobic methyl groups in the blue-emitting complexes but theeffect on improving the lifetimes of the LECs was limited [18 27 29ndash31 50 51]Tordera et al revealed that fluorine substitution in the complexes decreased thestability of the green LECs [6 17] but it appeared that avoiding fluorine substi-tution was not adequate to ensure good stability for blue or blue-green LECs Inaddition complexes with poor electrochemical stability ie irreversible oxidation
8 Recent Advances on Blue-Emitting Iridium(III) Complexes hellip 231
or reduction gave limited lifetimes for the LECs Thus both intrinsic and elec-trochemical stability should be considered for the complexes to improve the life-times of the LECs
Other factors such as the category and the content of the ionic liquid doped inthe emissive layer and the operation condition exerted significant influence on thestability of the blue or blue-green LECs Doping the ionic liquid or increasing thedriving voltages reduced the lifetime of the LECs although they shortened theturn-on time ie there was a trade-off between the lifetime and the turn-on time [34] Because relatively high driving voltages or currents are required for relativelyhigh brightness from the blue or blue-green LECs it would be even more difficult toenhance the lifetimes of the blue or blue-green LECs for solid-state lightingapplications It has been shown that pulsed current driving is superior over theconstant voltage driving for improving the overall performances of LECs includingresponse time efficiency and stability [21]
832 Challenges and the Future
As summarized in Sect 831 good color purity high efficiency and relativelyshort response time have been achieved for individual blue or blue-green LECsHowever one LEC with good color purity high efficiency and brightness as wellas short turn-on timesbquo has not been achieved yet Moreover all the blue orblue-green LECs exhibited limited lifetimes These challenges should be overcomethrough the development of high-performance blue-emitting materials and theoptimization of the devices
1 Materials
Through previous works (Table 81) requirements for high-performance ioniciridium(III) complexes for LECs can be summarized as
(i) Good color purity sufficient for the generation of white light This demandssmall red-shifts for the PL from solution to concentrated films along with xyCIE color coordinates of the EL around or better than 020040 and goodchemical or morphological stability of the complexes under electrical exci-tation for good stability of the EL color upon the continuous operation of theLECs
(ii) High luminescent efficiency in both solution and films This requires tailoredexcited-states such as large energy differences between the 3MC and theemitting triplet states and suppressed concentration-quenching in the con-centrated films
(iii) Good intrinsic stability including the chemical and electrochemical stabilityof the ligands and the complexes under the electrical excitation Thisdemands no C-F or other fragile chemical bonds on the ligands reversibleoxidation and reduction in cyclic voltammetry and rigid coordination bonds
232 L He
2 Devices
Although outstanding blue-emitting iridium(III) complexes establish the basis forblue or blue-green LECs their optimization is equally important to enhance theoverall device performances For example many factors such as (i) the use of ionicliquids andor carrier-trapper dopants in the emissive layers as well as the drivingmodes for the operation of LECs influence significantly the performances of theblue or blue-green LECs
Recently Pertegaacutes et al demonstrated a host-guest configuration for theblue-green LECs incorporating iridium(III) complexes [63] The device structurewas ITOPEDOTPSS (80 nm)NMS25 SPPO13 10 FIrpic (80 nm)Al whereNMS25 was a cationic carbazole-containing hole-transporter SPPO13 was a neu-tral electron-transporter and FIrpic was a blue-green iridium emitter This host-guestLEC has characteristics of both OLEDs and LECs In the device the ionic hostrather than the iridium(III) complex transport the electrons and holes and theiridium(III) complex as a phosphorescent dopant emit the blue-green light Thehost-guest configuration is promising for developing high-performance blue orblue-green LECs because it suppresses the concentration-quenching of thelong-lived triplets releases the complexes from carrier-transporting which thecomplexes may not be good at and reduces the use of the iridium(III) complexesand thus the cost for the device production It is believed that withhigh-performance blue-emitting materials together with optimization of the devi-ces blue or blue-green LECs suitable for applications will come out in the nearfuture
Acknowledgements This work was supported by the National Natural Science Foundation ofChina (Grant No 51403240)
References
1 JD Slinker AA Gorodetsky MS Lowry JJ Wang S Parker R Rohl S BernhardGG Malliaras J Am Chem Soc 126 2763 (2004)
2 JD Slinker J Rivnay JS Moskowitz JB Parker S Bernhard HD AbrunaGG Malliaras J Mater Chem 17 2976 (2007)
3 T Hu L He L Duan Y Qiu J Mater Chem 22 4206 (2012)4 RD Costa E Orti HJ Bolink F Monti G Accorsi N Armaroli Angew Chem Int Edit
51 8178 (2012)5 HJ Bolink E Coronado RD Costa E Orti M Sessolo S Graber K Doyle
M Neuburger CE Housecroft EC Constable Adv Mater 20 3910 (2008)6 D Tordera M Delgado E Orti HJ Bolink J Frey MK Nazeeruddin E Baranoff Chem
Mater 24 1896 (2012)7 KA King RJ Watts J Am Chem Soc 109 1589 (1987)8 F Neve A Crispini S Campagna S Serroni Inorg Chem 38 2250 (1999)9 RD Costa E Orti HJ Bolink S Graber S Schaffner M Neuburger CE Housecroft
EC Constable Adv Funct Mater 19 3456 (2009)10 S Ladouceur E Zysman-Colman Eur J Inorg Chem 2985 (2013)
8 Recent Advances on Blue-Emitting Iridium(III) Complexes hellip 233
11 JD Slinker CY Koh GG Malliaras MS Lowry S Bernhard Appl Phys Lett 86173506 (2005)
12 MS Lowry JI Goldsmith JD Slinker R Rohl RA Pascal GG Malliaras S BernhardChem Mater 17 5712 (2005)
13 HJ Bolink L Cappelli S Cheylan E Coronado RD Costa N LardiesMK Nazeeruddin E Orti J Mater Chem 17 5032 (2007)
14 MS Lowry WR Hudson RA Pascal AS Bernhard J Am Chem Soc 126 14129(2004)
15 HJ Bolink E Coronado RD Costa N Lardies E Orti Inorg Chem 47 9149 (2008)16 S Schmidbauer A Hohenleutner B Konig Adv Mater 25 2114 (2013)17 D Tordera JJ Serrano-Perez A Pertegas E Orti HJ Bolink E Baranoff
MK Nazeeruddin J Frey Chem Mater 25 3391 (2013)18 D Tordera AM Buenzli A Pertegas JM Junquera-Hernandez EC Constable
JA Zampese CE Housecroft E Orti HJ Bolink Chem Eur J 19 8597 (2013)19 EC Constable CD Ertl CE Housecroft JA Zampese Dalton Trans 43 5343 (2014)20 CD Ertl L Gil-Escrig J Cerdaacute A Pertegaacutes HJ Bolink JM Junquera-Hernaacutendez
A Prescimone M Neuburger EC Constable E Ortiacute CE Housecroft Dalton Trans 4511668 (2016)
21 NM Shavaleev R Scopelliti M Graetzel MK Nazeeruddin A PertegasC Roldan-Carmona D Tordera HJ Bolink J Mater Chem C 1 2241 (2013)
22 NM Shavaleev G Xie S Varghese DB Cordes AMZ Slawin C Momblona E OrtiHJ Bolink IDW Samuel E Zysman-Colman Inorg Chem 54 5907 (2015)
23 S Evariste M Sandroni TW Rees C Roldan-Carmona L Gil-Escrig HJ BolinkE Baranoff E Zysman-Colman J Mater Chem C 2 5793 (2014)
24 MK Nazeeruddin RT Wegh Z Zhou C Klein Q Wang F De Angelis S FantacciM Gratzel Inorg Chem 45 9245 (2006)
25 FD Angelis S Fantacci N Evans C Klein Inorg Chem 46 5989 (2007)26 L He L Duan J Qiao RJ Wang P Wei LD Wang Y Qiu Adv Funct Mater 18 2123
(2008)27 CD Sunesh K Shanmugasundaram MS Subeesh RK Chitumalla J Jang Y Choe
ACS Appl Mater Interfaces 7 7741 (2015)28 NM Shavaleev R Scopelliti M Graumltzel MK Nazeeruddin Inorg Chim Acta 404 210
(2013)29 CD Sunesh MS Subeesh K Shanmugasundaram RK Chitumalla J Jang Y Choe
Dyes Pigments 128 190 (2016)30 CD Ertl J Cerda JM Junquera-Hernandez A Pertegas HJ Bolink EC Constable
M Neuburger E Orti CE Housecroft Rsc Advances 5 42815 (2015)31 L He L Duan J Qiao D Zhang L Wang Y Qiu Chem Commun 47 6467 (2011)32 L He J Qiao L Duan GF Dong DQ Zhang LD Wang Y Qiu Adv Funct Mater 19
2950 (2009)33 P Pla JM Junquera-Hernandez HJ Bolink E Orti Dalton Trans 44 8497 (2015)34 CD Sunesh G Mathai Y Choe Org Electron 15 667 (2014)35 M Mydlak C Bizzarri D Hartmann W Sarfert G Schmid L De Cola Adv Funct Mater
20 1812 (2010)36 B Chen Y Li W Yang W Luo HB Wu Org Electron 12 766 (2011)37 J Li PI Djurovich BD Alleyne M Yousufuddin NN Ho JC Thomas JC Peters
R Bau ME Thompson Inorg Chem 44 1713 (2005)38 MK Nazeeruddin R Humphry-Baker D Berner S Rivier L Zuppiroli M Graumletzel
J Am Chem Soc 125 8790 (2003)39 H-F Chen C Wu M-C Kuo ME Thompson K-T Wong J Mater Chem 22 9556
(2012)40 F Dumura Y Yuskevitch G Wantz CR Mayer D Bertin D Gigmes Synth Met 177
100 (2013)
234 L He
41 CH Yang J Beltran V Lemaur J Cornil D Hartmann W Sarfert R FrohlichC Bizzarri L De Cola Inorg Chem 49 9891 (2010)
42 F Kessler RD Costa DD Censo R Scopelliti E Ortiacute HJ Bolink S Meier W SarfertM Graumltzel MK Nazeeruddin E Baranoff Dalton Trans 41 180 (2012)
43 SB Meier W Sarfert J Junquera-Hernaacutendez M Delgado D Tordera E Ortiacute HJ BolinkF Kessler R Scopelliti M Graumltzel MK Nazeeruddin E Baranoff J Mater Chem C 1 58(2013)
44 N Darmawan CH Yang M Mauro M Raynal S Heun JY Pan H BuchholzP Braunstein L De Cola Inorg Chem 52 10756 (2013)
45 DR Martir AK Bansal VD Mascio DB Cordes AF Henwood AMZ SlawinPCJ Kamer L Martiacutenez-Sarti A Pertegaacutes HJ Bolink IDW SamuelE Zysman-Colman Inorg Chem Front 3 218 (2016)
46 AB Tamayo S Garon T Sajoto PI Djurovich IM Tsyba R Bau ME ThompsonInorg Chem 44 8723 (2005)
47 C-T Liao H-F Chen H-C Su K-T Wong J Mater Chem 21 17855 (2011)48 J-S Lu H-F Chen J-C Kuo R Sun C-Y Cheng Y-S Yeh H-C Su K-T Wong
J Mater Chem C 3 2802 (2015)49 HC Su HF Chen FC Fang CC Liu CC Wu KT Wong YH Liu SM Peng J Am
Chem Soc 130 3413 (2008)50 H-F Chen W-Y Hung S-W Chen T-C Wang S-W Lin S-H Chou C-T Liao
H-C Su H-A Pan P-T Chou Y-H Liu K-T Wong Inorg Chem 51 12114 (2012)51 E Baranoff HJ Bolink EC Constable M Delgado D Haumlussinger CE Housecroft
MK Nazeeruddin M Neuburger E Ortiacute GE Schneider D Tordera R WalliserJA Zampese Dalton Trans 42 1073 (2013)
52 L He LA Duan JA Qiao GF Dong LD Wang Y Qiu Chem Mater 22 3535 (2010)53 G-G Shan H-B Li H-T Cao D-X Zhu P Li Z-M Su Y Liao Chem Commun 48
2000 (2012)54 J Wu F Li Q Zeng C Nie PC Ooi T Guo G Shan Z Su Org Electron 28 314 (2016)55 B Beyer C Ulbricht D Escudero C Friebe A Winter L Gonzaacutelez US Schubert
Organometallics 28 5478 (2009)56 NM Shavaleev R Scopelliti M Graetzel MK Nazeeruddin Inorg Chim Acta 388 84
(2012)57 S Ladouceur D Fortin E Zysman-Colman Inorg Chem 50 11514 (2011)58 JM Fernaacutendez-Hernaacutendez S Ladouceur Y Shen A Iordache X Wang L Donato
S Gallagher-Duval MDA Villa JD Slinker LD Cola E Zysman-Colman J MaterChem C 1 7440 (2013)
59 F Monti A Baschieri I Gualandi JJ Serrano-Perez JM Junquera-Hernandez D TonelliA Mazzanti S Muzzioli S Stagni C Roldan-Carmona A Pertegas HJ Bolink E OrtiL Sambri N Armaroli Inorg Chem 53 7709 (2014)
60 L He Z Wang C Yang L Duan R Tang X Song C Pan Dyes Pigments 13 76 (2016)61 Z Wang L He L Duan J Yan R Tang C Pan X Song Dalton Trans 44 15914 (2015)62 HJ Bolink L Cappelli E Coronado A Parham P Stossel Chem Mater 18 2778 (2006)63 A Pertegaacutes NM Shavaleev D Tordera E Ortiacute MK Nazeeruddin HJ Bolink J Mater
Chem C 2 1605 (2014)
8 Recent Advances on Blue-Emitting Iridium(III) Complexes hellip 235
Chapter 9Thermally Activated Delayed FluorescenceEmitters in Light-EmittingElectrochemical Cells
Michael Yin Wong and Eli Zysman-Colman
Abstract Thermally activated delayed fluorescence (TADF) represents a verypromising singlet harvesting mechanism that permits harvesting of both singlet andtriplet excitons in electroluminescent devices In this chapter the operating prin-ciple of TADF mechanism is introduced Two major classes of TADF emittersemployed in light-emitting electrochemical cell (LEC) devices small moleculeorganic compounds and copper(I) complexes are discussed in the context of theiroptoelectronic properties and LEC device performance metrics A critical outlookfor each class of emitters is also provided
Keywords Thermally activated delayed fluorescence Small molecules Copper(I)complexes Light-emitting electrochemical cells Lighting technologies
91 Introduction
Efficiency optimization of light-emitting devices is crucial to their cost effectivenessand hence marketability One key aspect to enhancing device efficiency is excitonmanagement Tremendous effort has been devoted to developing strategies to utilizeboth singlet and triplet excitons under device operation with particular attentionpaid to recruiting triplet excitons as these constitute statistically 75 of totalinjected charges [1] In order to appreciate emitter development in light-emittingelectrochemical cells (LEC) it is instructive to review the evolution of emitters inorganic light-emitting diodes (OLED) since this related electroluminescencetechnology is more mature Historically the first generation of emitters employed inOLEDs was fluorescent in nature which translated to only emissive singlets beingharvested to produce light The maximum internal quantum efficiency (IQE) ofthese devices was hence 25 Assuming Lambertian light emission only around
MY Wong E Zysman-Colman (amp)Organic Semiconductor Centre EaStCHEM School of ChemistryUniversity of St Andrews St Andrews KY16 9ST Fife UKe-mail elizysman-colmanst-andrewsacuk
copy Springer International Publishing AG 2017RD Costa (ed) Light-Emitting Electrochemical CellsDOI 101007978-3-319-58613-7_9
237
20 of the generated photons are able to escape the device and the maximumexternal quantum efficiency (EQE) is merely 5 [2] In order to address theshortfall in EQE phosphorescent OLEDs were developed that were based onorganometallic complexes the most popular class of which were those that con-tained iridium(III) The key merit of these complexes is their capacity to access tothe triplet excited state as a result of the increased intersystem-crossing (ISC) ratesdue to the large spin-orbit coupling constant of the heavy metal ion [1 3 4] Usingiridium(III) emitters 100 IQE has been achieved [5] However the drawbacks totheir use are the cost and more importantly the scarcity of these metals on EarthSince the first report in 2009 of their use in OLEDs [6] thermally activated delayedfluorescence (TADF) emitters have attracted intense attention from both academiaand industry [7ndash9] TADF emitters recruit both singlet and triplet excitons and theycan therefore achieve comparable efficiencies to phosphorescent emitters Thecentral feature of these emitters is the very small energy gap between the lowestsinglet excited state and the lowest triplet excited state (DEST typically lt 01 eV)Under these conditions triplet excitons can be thermally upconverted to emissivesinglet excitons by reverse intersystem-crossing (RISC) The singlet excitonsformed as a result of RISC then emit light by delayed fluorescence [10 11]Currently TADF OLED materials can be generally categorized into two classespurely organic compounds [7ndash9] and copper(I) complexes [12ndash14] While emittersfrom these two classes all have small DEST in common one important distinctionbetween them is the coexistence of TADF and phosphorescence in Cu(I) emitters[15] due to the larger spin-orbit coupling constant of the central metal (f 857 cmminus1)that effectively mixes the singlet and triplet excited states On the other hand onlyfluorescence is observed for purely organic TADF emitters because they containonly lighter elements such as carbon and nitrogen
The development of LECemitters has taken roughly the same pathAs summarizedin Chap 1 the first LEC device reported by Pei et al in 1995 was based on organicfluorescent polymer emitters where poly[2-(36-dioxaheptyloxy-14-phenylene](DOHO-PPP) poly(14-phenylene vinylene) (PPV) and poly[5-(2prime-ethylhexyloxy)-2-methoxy-14-phenylene vinylene] (MEH-PPV) were employed as blue green andorange emitters respectively [16] where they were applied together with lithiumtriflate (LiOTf) and poly(ethylene oxide) (PEO) as the ionic additive and ion con-ducting polymer respectively The former serves to create doping regions near theelectrode to facilitate charge injection while the latter provides a medium for ionicmovement (seeChaps 1 2 and 3) Themost popular class of emitters for LECs to dateare based on phosphorescent ionic transition metal complexes (iTMC) the firstexample of which was reported in 1996 by Maness et al [17] who used poly[Ru(vbpy)3][PF6]2 (vbpy = 4-vinyl-4prime-methyl-22prime-bipyridine) as the emitter Howeverthemost common iTMCs as emitters inLECs are based on iridium(III) complexes (seeChaps 4 5 7 and 8) the first example of which was reported by Slinker et al2004 [18ndash20] More recently cationic copper(I) complexes have drawn increasedattention as an alternative class of iTMCs to cationic iridium(III) complexes (seeChap 11) It is worth noting that copper(I) complexes can emit either by phospho-rescence TADF or both [21ndash23] Unlike OLEDs where small molecule organic
238 MY Wong and E Zysman-Colman
compounds are an important part of the emitter landscape only recently have organicemitters for LECs drawn increased attention the majority of these are purelyfluorescent in nature (see Chap 12) In this chapter we describe the potential of bothcharged small molecule (SM) organic and copper(I) complexes TADF emitters aseconomical replacements for the current state-of-the-art iridium(III) emitters As suchthis chapter constitutes a comprehensive introduction to the next Chaps 10 11 and 12focused on the state-of-the-art exciplex-like emission mechanism copper(I) com-plexes and SM-based LECs
92 Photoluminescence Mechanism FluorescencePhosphorescence and TADF
When a molecule is photo-excited a singlet excited state is first formed Afterinternal conversion to the lowest singlet excited state radiative relaxation to theground state becomes possible which is termed as fluorescence This is typicallyobserved in purely organic emitters that have only light atoms such as carbonhydrogen oxygen and nitrogen On the other hand organometallic complexeswhich have a heavy central metal such as iridium and platinum have highspin-orbit coupling constants (f) that permit a strong mixing between singlet andtriplet states In this manner intersystem-crossing (ISC) from the singlet excitedstate to the triplet excited state becomes feasible Radiative decay from thenow-accessed triplet excited state is termed phosphorescence In TADF emittersthe energy gap (DEST) between the lowest excited singlet (S1) and triplet (T1)excited states is small (typically lt 01 eV) Under these conditions the tripletexcitons can be thermally upconverted to singlet excitons by reverseintersystem-crossing (RISC) Assuming that the lowest energy transition is domi-nated by HOMO to LUMO DEST is related to exchange integral J [24 25]
DEST frac14 Es ET frac14 2J eth91THORN
where ES and ET represent the energies of the singlet and triplet excited statesrespectively and J describes the electron density overlap between the HOMO andLUMO defined in (92)
Jfrac14ZZ
HOMO r1eth THORNLUMO r2eth THORN 1r2 r1j jHOMO r2eth THORNLUMO r1eth THORNdr1dr2 eth92THORN
where HOMO and LUMO are the spatial distributions of HOMO and LUMOrespectively and r1 and r2 are position vectors It follows that reducing the overlapbetween HOMO and LUMO decreases the exchange integral (J) and hence DEST
In TADF material DEST is essential to the success of singlet harvesting becauseit governs the rate of RISC according based on a Boltzmann distribution [7]
9 Thermally Activated Delayed Fluorescence Emitters hellip 239
kRISC expDEST
kBT
eth93THORN
where kRISC is the rate constant of RISC kB is Boltzmannrsquos constant and T is thetemperature The singlet states formed as a consequence of RISC emit byfluorescence which is termed ldquodelayed fluorescencerdquo because the RISC process ismuch slower (typically on the order of ls to ms) compared with direct fluorescencewhich is termed ldquoprompt fluorescencerdquo and occurs on the other of nanoseconds Asa result triplet excitons can be productively utilized by transforming them intosinglet excitons While this RISC process is common to both purely organic andcopper(I) TADF emitters there is one distinct difference between the two In purelyorganic TADF emitters the triplet states do not emit by phosphorescence Howeverthe central copper atom has a moderately strong spin-orbit coupling constant (f857 cmminus1) and thus both TADF and phosphorescence are available as radiativedecay channels reflected in the different contributions to the delayed component ofemission lifetime (sd) In purely organic emitter sd is defined as in (94) [26]
1sd
frac14 kTnr thorn 1 kISCksr thorn ksnr thorn kISC
kRISC eth94THORN
where ksr and ksnr are the radiative and non-radiative decay rate constants of decaychannels from the S1 state respectively and kISC and kRISC are the ISC (S1 T1)and RISC (T1 S1) rate constants respectively On the other hand the overalldecay lifetime s in copper(I) TADF complexes is shown in (95) [13]
sfrac143 + exp DEST
kBT
3
sethT1THORN thorn1
sethS1THORN
h iexp DEST
kBT
eth95THORN
where s(T1) and s(S1) are the decay lifetimes of the triplet and singlet statesrespectively and kB and T are Boltzmann constant and temperature respectivelyThe distinguishing feature between Eqs (94) and (95) is the presence of an extraterm in Eq (95) that reflects the fact that in copper(I) TADF system the tripletstate is also emissive
93 SM-Based LECs
931 State-of-the-Art of Blue-Emitting SM-Based LECwithout TADF Mechanism
The most common emissive core studied for SM-based LECs is that of an oligo-fluorene as exemplified by 1 (Fig 91) which is a deep-blue fluorophore with
240 MY Wong and E Zysman-Colman
kem = 393 nm in dichloromethane (DCM) solution Chen et al [27] prepared anionic terfluorene emitter (1) that contains tethered charged imidazolium groups Theemitter gave deep-blue emission at 418 nm as a neat film with excellent photolu-minescence quantum yield () of 076 The presence of the long imidazoliumpendant groups in the emitter inhibits aggregation in the film regardless of thepresence or absence of small quantities of [BMIM][PF6] ionic liquid which istypically added to improve charge transport in the film [28] The best LEC device(ITOPEDOTPSS (30 nm)1[BMIM][PF6] (ww = 91 ca 200 nm)Al (100 nm))was achieved with an emissive layer constituted of 10 wt [BMIM][PF6] in 1The LEC operating at constant voltage showed a deep-blue emission color (kEL423 nm) with xy CIE color coordinates of 015012 and an EQE of 114 sig-nificantly bluer that LECs employing cationic iridium(III) complexes [20 29]Despite the addition of ionic liquid to the emissive layer the turn-on time for thisdevice was 51 min (defined as the time taken to reach maximum brightness seeChap 1) while the lifetime t12 defined as the time taken to reach half of themaximum brightness see Chap 1) was a 427 min Devices without [BMIM][PF6]indeed performed similarly but showed expectedly significantly longer turn-ontimes The photophysical and electrochemical properties of all emitters in thissection as well as their device metrics are summarized in Tables 91 and 92respectively
The same group then reported a similar polycationic bifluorene (2 Fig 92)system this time containing four tethered methylimidazolium groups [30] Theshortened effective conjugative length of the molecule resulted in a furtherblue-shifted emission into the ultraviolet (UV) region (kem = 373 nm in MeCN)The film morphology was found to improve when poly(methyl methacrylate)(PMMA) was dispersed at 10 wt doping in a 2 film Neither emission energy norphotoluminescence quantum yield was impacted by the use of doped films(kem = 383 nm = 41 as a 90 wt film in PMMA) As a result of improvedfilm morphology and employing a thick emissive layer of 350 nm the deviceperformance could be greatly enhanced when 2 was dispersed at very high con-centrations (90 wt) in a PMMA matrix Of the devices fabricated the one (ITO
N N
N N
1
2+
PF6-
2
Fig 91 Chemical structureof 1
9 Thermally Activated Delayed Fluorescence Emitters hellip 241
Tab
le91
Summaryof
device
performance
ofSM
-based
LECs
Emitter
Devicestructure
Von
(V)
t max
(min)
EL m
ax
(nm)
CIE
(xy)
EQE
()
PE (lmW
)Eff m
ax
(cdA)
t 12
(min)
Ref
2ITO
(120
nm)PE
DOTPSS
(30nm
)em
itterPMMA
(ww
=91
350nm
)Ag(100
nm)
minus67
388
(016010)
106
minusminus
97
[30]
1ITOPEDOTPSS
(30nm
)em
itter[BMIM
][PF
6](ww
=91
ca
200nm
)Al(100
nm)
32
5142
3(015012)
114
124
minus42
7[27]
4ITOPEDOTPSS
(80nm
)neat
emitter
(100
nm)Al
51
minus43
2(015009)
minusminus
015
minus[32]
545
minus43
4(016010)
minusminus
014
minus
3ITOPEDOTPSS
(40nm
)neat
emitter
orem
itter[BMIM
][PF
6](ww
=91)
(75nm
)Ag(100
nm)
minusminus
454
(016022)
minusminus
014
minus[31]
20ITOPEDOTPSS
(80nm
)neat
emitterAl
minusminus
470
(021025)
001
minusminus
minus[46]
9ITOPEDOTPSS
neatem
itterAl(100
nm)
87
minus48
4(018027)
minusminus
019
minus[36]
1058
minus48
7(019031)
minusminus
02
minus
8ITOPEDOTPSS
emitterPEOLiTf(www
=1010185)Al
(100
nm)
65
minus49
1(019036)
minusminus
038
minus
12ITOPEDOTPSS
(60-90
nm)ihpy
pnibp
bn(ww
=1006
)Al
(100
nm)
7minus
496
(023037)
minusminus
135
minus[37]
7ITOPEDOTPSS
emitterPEOLiTf(www
=1010185)Al
(100
nm)
85
minus50
3(021039)
minusminus
029
minus[36]
6ITOPEDOTPSS
(80nm
)em
itterPEOLiTf(w
ww
=1010185
90nm
)Al(100
nm)
43
minus52
1(038049)
minusminus
minusminus
[33]
17ITOPEDOTPSS
neatem
itterAl
40
minus53
0(032058)
minusminus
minusminus
[40]
19ITOPEDOTPSS
(80nm
)neat
emitter
(100
nm)Al(70nm
)
3(pulsed)
minus53
8(035057)
039
07
minusminus
[44]
1314
ITO(120
nm)PE
DOTPSS
(30nm
)1314(ww
=13
200nm
)Ag(100
nm)
minus12
155
0minus
304
1029
minus10
0[38]
Dye
HITOPEDOTPSS
(80nm
)neat
emitter
(80nm
)Al
3 (pulsed)
minusminus
minuslt0
001
minusminus
minus[47]
NoteThe
term
sVont
onE
QEP
EE
ff maxt 12B
maxrefer
toturn-onvoltageturn-on
time
external
quantum
efficiencyp
ower
efficiencyefficacylifetim
emaxim
umluminancemaxim
umefficiencyrespectiv
ely
The
definitio
nof
each
term
isprov
ided
inChap
1
242 MY Wong and E Zysman-Colman
(120 nm)PEDOTPSS (30 nm)2PMMA (ww = 91350 nm)Ag (100 nm)) withthe highest EQE at 106 with a turn-on time of 67 min (constant voltage oper-ation at 46 V) in the presence of the PMMA and dropped sharply to 014 for the
Table 92 Summary of photophysical and electrochemical properties of organic emitters used inLECs
Emitter Solution kPL(nm)
SolutionUPL
Solid state kPL(nm)
Solid stateUPL
HOMO(eV)
LUMO(eV)
Ref
2 373 (MeCN) 1 385 (neat) 045 584 214 [30]
1 393 (DCM) 1 418 (neat) 076 567 211 [27]
4 389 (DCM) 095 411 (neat) minus 558 234 [32]
5 390 (DCM) 098 407 (neat) minus 555 231
3 420 (MeCN) 077 454 (emittinglayer)
minus 564 254 [31]
21 459 (MeCN) 064 428 (neat) 061 552 527 [46]
20 463 (MeCN) 055 440 (neat) 044 585 550
9 450 (THF) 065 458 (neat) 035 minus minus [36]
10 453 (THF) 064 463 (neat) 036 minus minus
8 445 (THF) 065 459 (neat) 031 minus minus
12 506 (MeCN) 0002 minus minus 534 275 [37]
7 445 (THF) 064 462 (neat) 030 minus minus [36]
6 458 (THF) 060 477 (neat) 030 53 24 [33]
17 507 (PhMe) 010 514 (neat) 003 503 211 [40]
19 572 (MeCN) 002 500 (10 wt inPMMA)
090 587 299 [44]
1314 minus minus 545 (neat film) minus 523 280 [38]
Dye H 600(MeCN)
minus minus lt001 54 26 [47]
2
N N
N N
NN
NN
PF6-
4
4+Fig 92 Chemical structureof 2
9 Thermally Activated Delayed Fluorescence Emitters hellip 243
device using neat 2 the thickness of this latter device was reduced to 250 nm Thereduction in performance was attributed to the amphiphilic nature of the 2 whichtended to form nanoaggregates PMMA helped to distribute the emitter in a uniformmanner to avoid current leakage The best device fabricated with 90 wt 2 mixedwith 10 wt PMMA gave an electroluminescence maximum at 388 nm and xyCIE color coordinates at 016010 albeit the very short lifetime of 97 min Indeedall devices in this studied were very short-lived
Shanmugasundaram et al [31] reported an ionic fluorene emitter (3 Fig 93)with pyrene moieties installed at the terminal ends and two charged methylimida-zolium groups tethered from the 9-position of central fluorene 3 gave a deep-blueemission (kem = 420 nm) in MeCN with a high of 077 The emission of the neatfilm was red-shifted to 454 nm The presence of 10 wt [BMIM][PF6] does notalter the LEC device performance except for the more rapid turn-on times Bothdevices (ITOPEDOTPSS (40 nm)neat 3 or 3[BMIM][PF6] (ww = 91) (75 nm)Ag (100 nm)) in the report produced deep-blue emission at ca 454 nm and xy CIEcolor coordinates of ca 016022 with a luminance and current efficiency of1247 cdm2 and 014 cdA respectively No EQE and device lifetime data wereprovided in the report
The same group then reported two very similar emitters in which the pyrenemoieties were substituted for 6-alkoxy-substituted naphthalenes (4 5 Fig 94)[32] The photophysical profiles of 4 and 5 are essentially identical with emission at389 and 390 nm in DCM respectively and near-unity of 095 and 098respectively As a neat film the kem were slightly red-shifted at 411 and 407 nmrespectively no neat film data were provided The LEC devices (ITOPEDOTPSS (80 nm)neat 4 or 5 (100 nm)Al) expectedly gave near identical performancemetrics (the ITO and Al thicknesses were note reported) Taking the device with 4as the example the kEL was 432 nm and xy CIE color coordinates 015009 under
3
NN
NN
PF6-
2
2+Fig 93 Chemical structureof 3
244 MY Wong and E Zysman-Colman
constant voltage of 51 V Maximum luminance and current efficiencies of118 cdm2 and 014 cdA respectively show a poorly emissive device No EQE ordevice lifetime data were provided in the report
Subeesh et al [33] reported the use of the neutral emitter (6 Fig 95) in LECwhose phenanthroimidazole moiety has been previously reported to show excellentphotophysical properties and balanced charge transport [34 35] 6 showed anemission maximum of 458 nm and a of 060 in THF solution while as a neat filmkem = 477 nm and = 030 A yellowish-green emitting LEC device (ITOPEDOTPSS (80 nm)6PEOLiOTf (www = 1010185 90 nm)Al (100 nm)) showedkEL and xy CIE color coordinates of 521 nm and 038049 respectively withmaximum luminance of 125 cdm2 at 99 V The turn-on voltage (defined asbrightness reaching 1 cdm2 see Chap 1) of the device was 43 V but the turn-ontime efficiency and lifetime of the device were not reported
The authors then reported a series of charged and neutral PYPN derivatives(7 8 9 and 10) with different N-substituted alkyl groups (Fig 96) [36] Each ofthe emitters in Fig 96 has essentially the same emission energies (kem =458 minus 463 nm) and of 030ndash036 as neat thin films suggesting the length of thealkyl group charged or neutral has little effect on the chromophore The emission
NN
NN
R R
R = OMe R = OEt
45
PF6-
2
2+Fig 94 Chemical structuresof 4 and 5
HN
N
6
Fig 95 Chemical structureof 6
9 Thermally Activated Delayed Fluorescence Emitters hellip 245
in the film is slightly red-shifted and the reduced compared to measurements inTHF solution (kem = 445 minus 453 nm = 064 minus 065) Regardless of emittersky-blue emission was observed in the LEC (kEL = 484 minus 503 nm) with the bluestof the devices (ITOPEDOTPSSneat emitterAl (100 nm)) using 9 having xy CIEcolor coordinates of 018027 with a current efficiency of 019 cdA (the thicknessof the emissive layer was not reported) The device with 10 showed a very similaremission as 9 with xy CIE color coordinates and current efficiency of 019031 and020 cdA respectively which suggests the length of the tether to the methylimi-dazolium group had only a modest effect on the device performance Importantlythe emissive layer employing the ionic emitters was as a neat film in the devicewhile the neutral emitters were mixed with PEO and LiOTf in a weight ratio of1010185
The same group then studied a host-guest approach in LEC devices and foundout that selection of host influences critically the device performances [37] Inparticular they compared the performances of 12 (Fig 97) in two hosts (11Fig 97 and 10 Fig 96) [36] The emitter 12 possesses a pushndashpull design pro-ducing an intramolecular charge-transfer (ICT) emission which is solvatochromic
N
N
7
N
N
8
N
N
9
N
N
10
NN
NN
PF6
-+
PF6
-+
Fig 96 Chemical structures of 7 8 9 and 10
246 MY Wong and E Zysman-Colman
as evidenced by the red-shift from 489 nm to 506 nm when the media changes fromtoluene to acetonitrile The dropped significantly from 030 to 0002 withincreasing solvent polarity The 11 and 10 hosts gave kPL at 409 and 461 nmrespectively and of 095 and 063 respectively in acetonitrile The best device(ITOPEDOTPSS (60ndash90 nm)1012 (ww = 1006)Al (100 nm)) gave a maxi-mum brightness of 5016 cd mminus2 at a current efficiency of 073 cdA that is amongthe highest brightness for a LEC device operated under constant voltage reported todate (the thickness of the emissive layer was not reported) On the other handdevices using 11 as the host gave much lower device efficiencies (002ndash004 cdA)no EQE device lifetime or turn-on time data was reported
Chen et al [38] reported the use of an exciplex system as the emitting layer inthe LECs This strategy enjoys the advantages of reversible oxidation and reduction
N
N
N
N
11
N
N
N
N
N
N
N
N
NS
N
12
PF6
-+
PF6
-2+
2
Fig 97 Chemical structures of 11 and 12
9 Thermally Activated Delayed Fluorescence Emitters hellip 247