A look inside white OLEDs

Europhysics News, Jul 2018

White organic light-emitting diodes (OLEDs) for lighting are finding their way to the market. These OLEDs have organic layers only a few nanometres thick. Improving their efficiency and stability requires nanometre-precision information about where the light is emitted. Reconstruction of the emission profile from the measured emission and molecular-scale modelling provide a unique nanoscale look inside white OLEDs.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:


A look inside white OLEDs

A LOOK INSIDE WHITE OLEDs hoorn - DOI: White organic light-emitting diodes (OLEDs) for lighting are finding their way to the market. These OLEDs have organic layers only a few nanometres thick. Improving their efficiency and stability requires nanometre-precision information about where the light is emitted. Reconstruction of the emission profile from the measured emission and molecular-scale modelling provide a unique nanoscale look inside white OLEDs. - b P.21: A transparent white OLED made by Philips c FIG. 1: Schematic of the studied multilayer white OLED stack. Given are the chemical structures of the used organic molecular semiconductors and dyes (apart from proprietary materials from the company Novaled), and the thickness and function of the layers. See Ref. [ 4 ] for more details. Ations of smart-phones. The breakthrough of lmost everyone nowadays is familiar with the high-quality OLED displays of recent genera OLED TVs seems a matter of time. Advantages of OLED displays over the ubiquitous liquid-crystal displays (LCDs) are evident. LCDs require a permanently emitting white back-light,illuminating an array of switchable liquid-crystal cell pixels with colour filters,which tune brightness and colour of the transmitted light. Hence, LCD technology is based on light absorption. In contrast, OLEDs actively emit light of the desired colour from pixels of organic semiconductors.This leads to higher contrast ratios (“black”is really black) and lower power consumption. OLEDs for lighting applications [1] are presently under development,not in the first place because of power efficiency – present-day commercial inorganic LEDs are still more efficient – but because they are unique in being ultrathin large-area light sources which can be flexible and transparent.Therefore,an OLED lamp needs no special luminaire: it is its own luminaire.And as luminaire losses are avoided,the efficiency at a system level is expected to be competitive over a wide range of applications.The required broad-spectrum white light could in principle be obtained from a Red-GreenBlue pixelated OLED display with all pixels switched on. However, the lateral structuring needed for this would be unnecessarily complicated and expensive.White-OLED research therefore focuses on vertical multilayer structures, pioneered in Japan [ 2 ].Light of different colours is emitted in different organic layers stacked on top of each other,together composing white light. It is important that a large fraction of the light actually leaves the structure and can be used for illumination. An excellent review of the efficiency of the light outcoupling from OLEDs was given two years ago in this journal by Brütting et al. [ 3 ].Here,we focus on the light generation,and in particular on the question where in these structures emission of the different colours takes place.There is an intimate relation between this question and the question of outcoupling,since the position where emission takes place directly determines the light outcoupling efficiency. A hybrid white OLED In recent work together with the Institut für Angewandte Photophysik (IAPP) of the Technische Universität Dresden and Philips Research Aachen, our groups studied within the European FP7 project AEVIOM (Advanced ExperimentallyValidated Integrated OLED Model) the multilayer white-OLED stack shown in Fig.1 [ 4 ].The OLED has been fabricated at the IAPP by thermal evaporation in ultra-high vacuum of the organic materials displayed in this figure. The structure is ideal for a fundamental study, because most of the used materials have been well characterized in literature and all relevant processes can be addressed. The generation of the primary colours in this OLED is based on a hybrid principle, used extensively nowadays in commercially available white OLEDs. Green and red light are generated in layers of a host organic semiconductor doped by green and red phosphorescent dyes.The use of this dyedoping for light generation in OLEDs has been pioneered by researchers from the US [ 5 ]. A heavy metal atom in such dye molecules (in this case iridium) opens up – by its strong spin-orbit coupling – a radiative decay pathway for excitons (Coulombically bound electron-hole pairs) with a triplet spin configuration (parallel spins of electron and hole) next to singlet excitons (antiparallel spins). Hence, almost all excitons formed in such layers decay under the emission of a photon.Hybrid OLEDs avoid the use of blue phosphorescent dyes – of which the stability is still an issue – by using instead blue-emitting molecules without a heavy metal atom, at which only the singlet excitons decay radiatively (fluorescence). This compromises the internal quantum efficiency, because the triplet excitons – 75% of all formed excitons – have no efficient radiative decay pathway and are thus wasted. Still, the power efficiency of today’s commercial hybrid white OLEDs is already a factor of three to four higher than that of incandescent light bulbs (~40-60 lm/W vs.~15 lm/W),combined with operational lifetimes exceeding 10,000 hours. Of crucial importance to the functioning of multilayer OLEDs is that excitons are generated at the right place by encounter of electrons and holes, and that their subsequent motion until the moment of radiative decay is precisely controlled. In the OLED of Fig. 1 an excitonblocking interlayer has been inserted in between the blue and green layer. This interlayer prevents the motion of singlet excitons from the blue to the green layer as well as that of triplet excitons from the green to the blue layer. These are unwanted energetically downward processes that have to be blocked. On the other hand, motion of excitons from the green to the red layer can take place because of the direct contact between these layers. This is a desired process, because it leads to the right colour balance in this OLED, as will become clear. Here, the link with the light-outcoupling problem comes Next to the control of the exciton motion, the control of into play. If it is known where in the OLED the emission the motion of electrons and holes is crucial. This control takes place,the external emission can be readily calculated occurs by using organic semiconductors with appropriate by optical microcavity modelling,using the known dielecenergy levels of electrons and holes; see Fig. 2. First, elec- tric properties of all layers in the stack and representing a trons and holes have to be injected from suitable electrodes radiatively decaying exciton as a classical dipole antenna (of which at least one has to be transparent, in this case [ 3 ].The totaloutcoupled radiation of the OLED is the sum indium-tin-oxide,ITO) into the organic layers.In the early of the outcoupled radiation of all dipoles in the stack.Howdays of OLEDs, researchers struggled with the difference ever, it is also possible to solve the inverse problem, that in energy between charges in the electrodes and in the is, to reconstruct the emission profile of all dipoles in the organic semiconductor,leading to large barriers for charge stack from knowledge of the total outcoupled radiation injection. This problem has practically disappeared by an field [ 7 ]. Solving such an inverse problem is similar to reinvention of the group at the IAPP. They introduced the constructing the three-dimensional (3D) inner structure use of highly n- and p-doped organic layers adjacent to of the earth from seismic signals (see the Feature by Jobert the electron- and hole-injecting contacts [ 6 ]. These layers in this issue), or the 3D structure of a patient’s body by behave as good electron and hole conductors, providing X-ray Computed Tomography (CT) scanning. Here, we an almost barrier-free contact with the electrodes. From make use of the strong dependence of the emission patthese doped layers the electrons and holes smoothly enter tern on the distance of the dipole emitter to the aluminium the electron- and hole-transporting layers, via which they cathode layer (see box). The information contained in the move to the inner layers of the stack.The electron and hole angle-, wavelength-, and polarization-dependent emitted energies of the organic semiconductors used in the trans- light intensity is sufficiently rich to determine the dipoleport layers are such that charge carriers of only one polarity emission profile of the primary colours with a nanometrecan enter these layers. This guarantees that electrons and scale precision.The result for the OLED of Fig.1 at a typical holes meet in the inner emissive layers of the stack and operating bias voltage is given in Fig. 3(a). The balance form excitons there. The OLED functions as an optical between emission of the primary colours,with a strong red microcavity, in which exciton formation close to metal- component, leads for this OLED to the emission of pleaslic electrodes must be avoided because this would lead to ant warm-white light. Resolution of the emission profile non-radiative decay.By carefully tuning the thicknesses of within the very thin (3 nm) green layer is just beyond the all layers, the outcoupling efficiency can be optimized [ 3 ]. limits of the approach. The profiles in the red and blue A special role is again played by the interlayer between the layer are on the scale of a few nanometres confined to the green and blue layer.This very thin (3 nm) layer should block interfaces with the green layer and interlayer, respectively. excitons, but allow passage of both electrons and holes, in order to guarantee exciton formation in both the green and blue layer. In order to achieve this, the interlayer consists of a mixture of an electron transporter and a hole transporter, where the electron energy of the former matches well to that of the blue fluorescent material, while the latter material is the same as the hole-transporting host in the green layer. Simulations at the molecular scale Unique insight in the interplay of the processes leading to the emission profile can be obtained from molecularscale simulations. Whereas inorganic LEDs are based on crystalline semiconductors, in which the electronic and excitonic wave functions are delocalized, OLEDs are based on amorphous materials, in which the electronic contact between the molecules is weak. The quantummechanical wave functions of electrons, holes, and excitons in such disordered materials are localized on the molecules. Their motion involves sudden hops from one c Reconstructed light-emission profile So, how does the OLED of Fig. 1 work? To answer this question, one would have to “look inside” the OLED, in order to find out where in the OLED the light is emitted. molecule to another by an incoherent phonon-assisted exciton motion is simulated within the green and red tunnelling process. This motion can ideally be simulat- layer, and, importantly, from the green to the red layer. ed by Monte Carlo methods, where at every time step a Excitons formed in the blue layer will stay there, because possible hop is chosen according to a randomly drawn of the adjacent exciton-blocking interlayer. Information number [ 8 ]. We have performed such simulations by about the radiative decay efficiencies of the blue fluomodelling the OLED stack as an array of hopping sites rescent and the green and red phosphorescent emitters, representing all the different molecules in the stack, in- determining the fraction of excitons that decay by emitcluding the dyes [ 4 ]. Electron traps occur in many or- ting a photon, is taken from experiments. This informaganic semiconductors. These are taken into account in tion is needed to predict the light-emission profile from the layers where they matter: the electron-transporting the simulations. Excitons formed in the interlayer are and blue fluorescent layers.The inter-site distance is taken assumed to be lost. to be 1 nm, which is the typical distance between the A first check of the validity of the simulations is the commolecules.All molecules are given an electron and a hole parison between the calculated and measured current in energy according to the energy-level scheme of Fig. 2, the OLED. At the operating voltage of 3.6 V the current with a random energy due to disorder added. The doped densities agree to within 25% [ 4 ], which is a gratifying electron- and hole-conducting layers are modelled as result in view of the rather drastic approximations and metallic-like contacts. The hopping rates are chosen to assumptions made. In Fig. 3(b) the simulated exciton reproduce available experimental information about the generation profile is given. We find that indeed almost mobility of electrons, holes, and excitons in each material all injected electrons and holes form excitons. As dein the stack.Coulomb interactions between all charges are sired, most excitons form in the emissive layers, with taken into account. Electrons and holes attracting each the majority of excitons (54%) formed in the blue layer. other by the Coulomb force form excitons. Subsequent Also, a considerable fraction of excitons is formed in the BOX 1: RECONSTRUCTION OF THE LIGHT-EMISSION PROFILE The light-emission profile in OLEDs can be determined with nanometre-scale resolution, far below the diffraction limit [ 7 ]. That is possible due to the interference of directly emitted light from a dipole and light emitted from its image dipole in the highly conducting aluminium cathode layer. The left figure shows the angular dependence of the emission from a randomly oriented green emitter in glass, at varying distances from the cathode, for the two polarization directions. Due to interference, the perpendicular emission intensity oscillates as function of distance to the cathode with a periodicity λ/(2n) ≅ 177 nm, where λ is the wavelength in vacuum (530 nm) and n is the refractive index of glass (1.5). In the distance range from 80 to 160 nm (approximately from the first maximum for the perpendicular emission, near λ/(4n), to the first minimum, near λ/(2n)) the emission profile depends for both polarizations strongly on the emission angle. Although the emission in actual OLEDs is influenced by the more complex layer structure, the dominant role of the metal cathode on the light emission pattern is preserved. In order to prevent loss of information due to internal reflection at the interface between the glass substrate and air, a glass hemisphere is used to extract also emission at large internal angles (figure on the right). Optical microcavity modelling and leastsquares fitting methods are used to obtain the emission profile in the OLED from the measured angle-, wavelength- and polarization-dependent emission. n . LEFT:Polardiagramsofthenormalizedsandp-polarizedemissionfromarandomlyorienteddipoleemitteringlassasafunctionofdistancetoanaluminiumlayer. RIGHT:measurement configuration.AglasshemisphereisputontopoftheOLEDtoextractalsolightwhichisinternallyemittedunderangleslargerthanthecriticalangleforfullinternalreflection,θcr ≈42°. 62% 19% 19% green layer (22%), while almost no excitons are formed in the red layer (3%). Fig. 3(b) reveals an important loss mechanism caused by excitons formed in the interlayer (21%). This leads to a somewhat suboptimal efficiency, as is indeed observed. After taking into account the excitonic motion and radiative decay efficiencies the simulated light-emission profile of Fig. 3(c) is obtained.We observe the same large component of emitted red light as found in the reconstruction of the experimental light-emission profile of Fig. 3(a), which is almost completely caused by transfer of excitons from the green to the red layer.Also the green and blue component of the simulated emission profile are in fair agreement with the reconstructed emission profile. Like the reconstructed profiles, the simulated profiles are confined to nanometre-scale regions close to the interfaces. The overall agreement between the reconstructed and simulated emission profiles is striking. A bright future The experimental and simulation tools discussed here provide a unique way to “look inside the OLED” and can be fruitfully used in the further development of white OLEDs, which show ever more complex layer structures. They could help in identifying what processes limit their efficiency and lifetime. An important current issue is to what extent interactions between two excitons and between excitons and charges lead to exciton quenching,and thereby to a loss of efficiency. It is found that at high voltages the efficiency of OLEDs drops (an effect called“roleoff ”), which is presumably caused by these interactions. We are presently investigating this issue using our tools. With all-phosphorescent OLEDs, power efficiencies beyond 100 lm/W have been demonstrated in laboratory experiments, even exceeding those of fluorescent lamps (~90 lm/W), while operating at a high luminance [ 9,10 ]. Novel concepts for efficient and stable organic light emitters are also emerging, such as organic emitters in which the energies of singlet and triplet excitons are almost identical [11]. These emitters allow a transformation of non-emissive triplet excitons into emissive singlet excitons, providing an alternative to phosphorescent emitters. Use of these emitters could lead to a radical redesign of OLEDs, to which our tools can also be applied. But whatever turn this will take, the future of OLEDs is bright. n About the authors Peter Bobbert (1960) studied and obtained his PhD (cum laude) in Theoretical Physics in Leiden. After a postdoc in Delft as Royal Dutch Academy of Sciences Fellow he joined Eindhoven University in 1991. He presently works on various aspects of the theory of Organic Electronics, among which OLED modelling. Reinder Coehoorn (1956) studied experimental physics in Amsterdam and obtained his PhD in Groningen. Since 1985, he is employed at Philips Research, working on permanent magnets,magnetic multilayers,and presently on OLEDs.He is part-time professor at the Eindhoven University,and an elected member of the Royal Netherlands Academy of Sciences. [1] For example Philips: www.lighting.philips.com/main/led/oled/ and Osram: www.osram.com/osram_com/news-and-knowledge/oled-home [2] J. Kido , M. Kimura , K. Nagai , Science 267 , 1332 ( 1995 ) [3] W. Brütting , J. Frischeisen , B.J. Scholz , T.D. Schmidt , Europhys. News 42 , 20 ( 2011 ) [4] M. Mesta , M. Carvelli , R.J. de Vries, H. van Eersel, J.J.M. van der Holst , M. Schober, M. Furno , B. Lüssem , K. Leo , P. Loebl , R. Coehoorn , P.A. Bobbert , Nature Mater. 12 , 662 ( 2013 ) [5] M.A. Baldo , D.F. O'Brien , Y. You , A. Shoustikov , S. Sibley , M.E. Thompson , S.R. Forrest , Nature 395 , 151 ( 1998 ) [6] K. Walzer , B. Maennig , M. Pfeiffer , K. Leo , Chem. Rev. 107 , 1233 ( 2007 ) [7] S.L.M. van Mensfoort , M. Carvelli , M. Megens , D. Wehenkel , M. Bartyzel , H. Greiner , R.A.J. Janssen , R. Coehoorn , Nature Photon. 4 , 329 ( 2010 ) ; M. Carvelli , R.A.J. Janssen , R. Coehoorn , J. Appl. Phys . 110 , 084512 ( 2011 ) [8] For a movie, see www .youtube.com/watch?v=n1qkzv_ Q7UU&feature=youtu .be [9] S. Reineke , F Lindner , G. Schwartz , N. Seidler , K. Walzer , B. Lüssem , K. Leo , Nature 459 , 234 ( 2009 ) [10] Y.-S. Park , S. Lee , K.-H. Kim , S.-Y. Kim, J.-H. Lee , and K.-J. Kim , Adv. Funct. Mat. ( 2013 ), DOI:10.1002/adfm.201300547 (advance online publication). [11] H. Uoyama , K , Goushi, K. Shizu , H. Nomura , C. Adachi , Nature 492 , 234 ( 2012 )

This is a preview of a remote PDF: https://www.europhysicsnews.org/articles/epn/pdf/2013/05/epn2013445p21.pdf

Peter Bobbert, Reinder Coehoorn. A look inside white OLEDs, Europhysics News, 21-25, DOI: 10.1051/epn/2013504