Triplet-triplet annihilation in highly efficient fluorescent organic light-emitting diodes: current state and future outlook.

Downloaded from http://rsta.royalsocietypublishing.org/ on May 21, 2015

rsta.royalsocietypublishing.org

Review Cite this article: Kondakov DY. 2015 Triplet–triplet annihilation in highly efficient fluorescent organic light-emitting diodes: current state and future outlook. Phil. Trans. R. Soc. A 373: 20140321. http://dx.doi.org/10.1098/rsta.2014.0321 Accepted: 5 February 2015 One contribution of 11 to a discussion meeting issue ‘Organic semiconductor spintronics: utilizing triplet excitons in organic electronics’. Subject Areas: solid-state physics Keywords: triplet–triplet annihilation, triplet fusion, upconversion, organic electroluminescence, quantum efficiency Author for correspondence: Denis Y. Kondakov e-mail: [email protected]

Triplet–triplet annihilation in highly efficient fluorescent organic light-emitting diodes: current state and future outlook Denis Y. Kondakov DuPont OLEDs, Wilmington, DE 19805, USA Studies of delayed electroluminescence in highly efficient fluorescent organic light-emitting diodes (OLEDs) of many dissimilar architectures indicate that the triplet–triplet annihilation (TTA) significantly increases yield of excited singlet states—emitting molecules in this type of device thereby contributes substantially to their efficiency. Towards the end of the 2000s, the essential role of TTA in realizing highly efficient fluorescent devices was widely recognized. Analysis of a diverse set of fluorescent OLEDs shows that high efficiencies are often correlated to TTA extents. It is therefore likely that it is the long-term empirical optimization of OLED efficiencies that has resulted in fortuitous emergence of TTA as a large and ubiquitous contributor to efficiency. TTA contributions as high as 20–30% are common in the state-of-the-art OLEDs, and even become dominant in special cases, where TTA is shown to substantially exceed the spin-statistical limit. The fundamental features of OLED efficiency enhancement via TTA— molecular structure-dependent contributions, current density-dependent intensities in practical devices and frequently observed antagonistic relationships between TTA extent and OLED lifetime—came to be understood over the course of the next few years. More recently, however, there was much less reported progress with respect to all-important quantitative details of the TTA mechanism. It should be emphasized that, to this day and despite the decades of work on improving blue phosphorescent OLEDs as well as the recent advent of thermally activated delayed fluorescence OLEDs, the majority of practical blue OLEDs still rely on TTA. Considering such practical importance

2015 The Author(s) Published by the Royal Society. All rights reserved.


Since their invention [1] in the late 1980s, organic light-emitting diodes (OLEDs) showed remarkable improvements in all characteristics related to their usability. Considering that the OLEDs can be broadly classified as energy conversion devices, or, more specifically, devices transforming electric power into light (unitless), the power efficiency of such a conversion is one of their primary characteristics, along with durability and emission colour. Operating voltage— one of the two primary constituents of power efficiency—is obviously just as important with respect to overall power efficiency as quantum efficiency, which is commonly defined as the unitless ratio of created or emitted photons to injected electrons. Currently, it is typical for the various state-of-the-art OLEDs to operate at voltages close to the thermodynamic limit, which is determined by the weighted average wavelength of emitted light. It is worth mentioning that, because the operating voltage is related to the current density, which itself is dictated by fairly subjective and ever-changing design considerations of target luminance of practical devices, closeness to the thermodynamic limit is also somewhat arbitrary. Nonetheless, considering steep relationships between voltage and current density typically observed in OLEDs, as well as a limited range of maximum luminances required for practical applications, operating voltages within a few volts of a thermodynamic limit are often considered to be ‘close’ to the theoretical limit. Furthermore, although minor reductions in operating voltages can be frequently achieved through specific device structure optimization involving, for example, transport and injection material selection, such improvements may be associated with some undesirable side effects, such as a decline in device durability [2–4]. Fine details aside, there is clearly a striking improvement in operating voltages of OLEDs when compared, for example, with approximately 6.5 V and approximately 14 V values at 20 mA cm−2 for the first examples of, respectively, small molecule-based [1] and polymer-based [5] devices, both emitting green-yellow light with an average photon energy slightly above 2 eV. In a similar manner, the efficiency of OLEDs has been greatly enhanced by more than two decades of intense research and development. From the early days of OLED development, the quantum efficiency was intended to serve as a consistent measure of how well injected carriers are converted into photons. However, in the absence of a readily accessible and reliable method of measuring photon generation rates within the active medium, quantum efficiencies are typically reported in a form of external quantum efficiency (EQE), which includes an outcoupling term— the ratio of photons emitted from the device to the total number of photons produced. A relatively high refractive index of the organic medium is responsible for the outcoupling term being substantially below unity—usually estimated to be in the 0.2–0.3 range. Because only an approximate value of the outcoupling term is usually known for a given device structure, efficiency comparisons between dissimilar devices should be interpreted with caution. There are also additional sources of uncertainty in commonly reported EQE values due to technical details of measurement and related assumptions. For example, EQE values are frequently obtained by measuring the photon flux in the normal direction only and extrapolating the total photon output based on the assumption of the Lambertian nature of an OLED radiator. However, the significant deviations from Lambertian behaviour are relatively common in OLEDs and, considering that a normal measurement samples only a minor fraction of light emitted by a device, there may be substantial errors in EQE values obtained in this manner. More appropriate methods to measure total light emitted by a device in question include using an integrating sphere directly or interpolating an experimental angle profile for the device in question.

.........................................................

1. Introduction

2

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 373: 20140321

of fluorescent blue OLEDs, the design of blue OLED-compatible materials capable of substantially exceeding the spin-statistical limit in TTA, elimination of the antagonistic relationship between TTA-related efficiency gains and lifetime losses, and designing devices with an extended range of current densities producing near-maximum TTA electroluminescence are the areas where future improvements would be most beneficial.


(1.1)

where ηext is the optical outcoupling, approximately related to the refractive index of the emissive medium n (ηext ∼ 0.5 n−2 ), φl is the luminescence quantum yield, ηr is the probability of formation of an emissive excited state in a recombination event, and γ is the charge-balance factor, which is essentially the fraction of total current that results in recombination. It is noteworthy that the φl term refers to an operating device and, as such, is decreased by various dynamic quenching processes due to the presence of charge carriers and electronically excited molecules. Depending on the multiplicity of the emissive state, OLEDs are normally classified into two types: the fluorescent devices, where the emissive state is excited singlet, and the phosphorescent devices, where the emissive state has triplet or higher multiplicity. Considering that spin statistics dictate a 3 : 1 ratio between triplet- and singlet-yielding recombination events, the ηr term is frequently assumed to be 0.25 for fluorescent devices and 1.0 for phosphorescent devices, which may be expected to result in 4× lower efficiencies of fluorescent devices on average. Such a putative shortcoming of fluorescent OLEDs may be expected to make them undesirable in practical applications. Nonetheless, due to the superior durability and emission colour they are still essential in many cases, particularly where a blue colour component is required.

2. Evolution of efficiencies of fluorescent organic light-emitting diodes Figure 1 depicts the evolution of efficiencies of fluorescent OLEDs since their invention. The efficiency gain is quite remarkable—more than an order of magnitude—and, because it has an appearance of a smooth mathematical function such as geometric progression, it may be (incorrectly) interpreted as an indication of monotonous incremental improvements in OLED materials and architectures without radical changes in the mechanism of operation. However, equation (1.1) predicts a maximum EQE of only approximately 5% in the fluorescent OLEDs [6]. Even taking a dramatically higher estimate for optical outcoupling, the maximum EQE is predicted to be below 10% in an ideal fluorescent OLED devoid of any significant loss processes. Hence, it became clear after the mid-2000s that the experimental observations quite significantly surpass even the most generous theoretical predictions. Considering that equation (1.1) works quite well for earlier OLEDs, there must be a radical change in the operating mechanism, negating one or more assumptions about its terms. And because the only term of equation (1.1) not already maximized in predicting the maximum EQE of fluorescent device predictions was ηr —the probability that an emissive excited state is formed in a recombination event—the experimental results strongly suggest that the fluorescent OLEDs may generate substantially more than 25% of emitting states—excited singlets—per recombination event. The recombination of charge carriers in OLEDs is a process involving two initially uncorrelated doublet states, which usually are spatially localized radical cations and radical anions in organic dielectric materials. Spin statistics dictate that the initial product of their encounter will have 25% probability of having singlet multiplicity and 75% probability of having triplet multiplicity. This ratio is the reason ηr was assumed to be 0.25 in the classical estimates of OLED efficiency. Aside from the theoretical considerations, the 3 : 1 ratio was also experimentally demonstrated by Baldo and co-workers [13] in an archetypal small molecule-based device. However, the transformation from the initial encounter to final products of recombination may not be a simple, spin-multiplicity-preserving process. By way of illustration, consider recombination between the radical cation of a common hole-transporting material NPB (N, N’diphenyl-N, N’-bis(1-naphthyl)-1,1 -biphenyl-4,4 -diamine) and the anion radical of a common hole injection material HAT-CN (2,3,6,7,10,11-hexacyano-1,4,5,8,9,12-hexaazatriphenylene). Because of the close alignment between the NPB highest occupied molecular orbital (HOMO)

.........................................................

EQE = ηext φl ηr γ ,

3


The first organic electroluminescent devices showed EQEs of the order of 1% and 0.05% for the small molecule-based and polymer-based devices, respectively [1,5]. To understand the magnitudes of these experimental values, it is instructive to treat EQE as a product of several separate terms [6],


4

20

...very high external electroluminescence quantum efficiency, of 19.3±1.5%, was achieved... (Adachi and co-workers [12])

EQE (%)

...significant boost in efficiency—up to 12% external quantum efficiency... (Novaled [3] ) ...we report high-efficiency fluorescent red OLEDs with 11.5% EQE... (Kodak [10])

10

...high external quantum efficiency of nearly 10%... (Kanno and co-workers [9]) improvement in the hole/electron recombination efficiency... high EQE... of 8.7%... (Liao et al. [8])

‘classical’ estimates of maximum efficiency (i.e. singlets only) 5 today, after the intense research effort, external QE of 3–4% can be obtained... (Kido & Iizumi [7])

0 1985

high external quantum efficiency (1% photon/electron)... achiveable... (Tang & VanSlyke [1])

1990

1995

2000

2005

2010

year

Figure 1. Evolution of the efficiencies of the fluorescent OLEDs [1,7–12]. (Online version in colour.)

and the HAT-CN lowest unoccupied molecular orbital (LUMO) energy levels, a heterojunction formed by these two materials is capable of dark generation of mobile charges in moderate electric fields, a process that is frequently used in practical OLEDs to generate holes. The reverse reaction is a typical recombination process, with one distinctive feature: the same close alignment of energy levels that allows dark charge generation also unavoidably precludes recombination from producing any electronically excited states of the molecules involved in a recombination event. In fact, the only electronic states energetically accessible as final products of such recombination are the ground singlet states. Therefore, the yield of singlets is 100% and the yield of the emissive excited state (ηr ) is 0%. Although this example neither explains high efficiencies nor suggests the possibility of higher-than-expected yields of the emissive singlets, it clearly demonstrates that the spin-statistical 3 : 1 ratio can be drastically perturbed, for example, by the energetic factor. In fact, the deviations resulting in an increased yield of singlet excited states have also been proposed to occur in polymer electroluminescence as early as 1999 [14]. Although these initial findings were supported by soon-followed experimental and theoretical work [15,16], more recent studies suggest that the excited singlet yield of recombination is unlikely to exceed 25%. For an in-depth summary of the fascinating evolution in understanding the recombination yield in polymeric materials the reader is directed to a recent review [17]. With 25% recombination yield of excited singlets, the only plausible explanation of substantially-above-the-estimate efficiencies of fluorescent devices requires some form of postrecombination conversion of triplet states into emissive singlet excited states. In principle, the possible mechanisms can be classified into two groups: (i) intersystem crossing, involving oneto-one conversion of triplets into excited singlets, thus summarily producing up to 100% excited singlets, and (ii) triplet–triplet annihilation (TTA; also known as triplet fusion), which requires a minimum of two triplet states to produce one singlet excited state. As first shown by Deaton et al. in 2010 [11], the former mechanism is indeed capable of producing fluorescent OLEDs with remarkably high efficiency (above 16% EQE), which is comparable to the efficiency of a typical phosphorescent OLED [11]. Later, Adachi and co-workers [12] demonstrated diverse fluorescent OLEDs with even higher efficiencies and coined the term TADF (thermally activated

.........................................................

15


...external quantum efficiency of 16.1%, demonstrating that triplet excitons can be harvested very efficiently through the delayed fluorescence channel... (Deaton et al. [11])


In contrast to the enhancement mechanisms outlined above, there is no fundamental reason why the TTA mechanism would require drastically altered structure and properties of emitting molecules or adjacent materials relative to normal OLED design rules, e.g. a high radiative rate and quantum yield of fluorescence. Hence, the fortuitous TTA processes could explain a substantial part of the curve depicted in figure 1. In fact, observations of delayed fluorescence in organic solids and the TTA mechanism as its plausible source predate the invention of OLEDs by several decades (for a review of an early work, see [19]). Hence it may be somewhat surprising that the first suggestion of the TTA-related enhancement of the efficiency of fluorescent OLEDs appeared only in 1998 [7]. The higher-than-expected EQE at unconventionally high current densities in doped Alq3 -based OLED led authors to propose the TTA as a source of the observed enhancement. A few years later Ganzorig & Fujihira [20] reported qualitatively similar enhancement also in an Alq3 -based device. Although the maximum efficiency was relatively low, presumably due to the absence of an efficient emitter, the substantial component of the luminance was found to scale quadratically with current density, suggesting a bimolecular process such as TTA. It is noteworthy that, although the authors do not point that out specifically, the persistence of a quadratic dependence at high current densities necessarily means that the putative TTA component is only a minor channel of consumption of triplets. Otherwise, the concentration of triplets would be expected to scale approximately as the square root of current density and the TTA contribution would scale approximately linearly with current density. The same year Sinha & Monkman [21] reported observations of the delayed fluorescence in electroluminescence of polymer-based devices. Based on the kinetics of the decay at different temperatures, the delay emission was interpreted as a manifestation of the TTA mechanism. Interestingly, the nearly-second-order decays even after a 50 µs delay suggest that TTA is a major pathway of triplet consumption. This implies that the intensity of the TTA-related emission must be a very substantial component of total electroluminescence. Unfortunately, no experimental data were provided to confirm this feature.

.........................................................

3. Early evidence of triplet–triplet annihilation in fluorescent organic light-emitting diodes

5


delayed fluorescence) for the mechanism. However, the TADF-based OLEDs require emitting molecules with relatively unusual photophysical properties, most particularly a small energy gap between the emissive singlet state and the triplet state populated by recombination. Because the emitters with a small singlet–triplet energy gap do not only readily undergo thermally activated upconversion, but also possess undesirably low oscillator strengths for the radiative transition between the excited singlet and ground states, the OLED emitters used in practical fluorescent devices prior to 2010 were unlikely to benefit significantly from the TADF mechanism. Another possible mechanism of one-to-one conversion of triplets into singlets was proposed by Baldo and co-workers [18]. Rather than employing molecules with a small energy gap between the lowest molecular excited states of triplet and singlet multiplicity, this approach exploited the possibility of an intersystem crossing step prior to the formation of molecular excited states. Under certain conditions, the charge-transfer (CT) states formed from localized radical ions— charge carriers—may have small singlet–triplet energy gaps, fast intersystem crossing rates, an excited singlet state below that of the respective triplet, and a faster rate for relaxation of the singlet CT state to the molecular excited singlet state than the analogous process in the triplet CT state. The potential advantage of this approach is that implementing a small energy gap in CT states may not require specially designed molecular structures. On the other hand, the effectiveness of such a conversion hinges on intersystem crossing outcompeting internal conversion. Therefore, a ‘mixing layer’ containing heavy atom molecules to enhance spin–orbit coupling and increase the intersystem crossing rate was proposed. Although some promising results were initially shown, there is no experimental evidence so far that this process may contribute significantly in efficient fluorescent OLEDs.


Taking advantage of a diverse set of readily available OLED devices in Eastman Kodak Company R&D, which pioneered and was actively developing this technology until 2010, we set out to examine the extent of TTA involvement in fluorescent OLEDs. Interestingly, we found that the delayed emission intensity was quite significant in the majority of available devices. There was also a fairly clear but noisy trend: on average, more efficient OLEDs were exhibiting a higher initial intensity of delayed emission. Apart from this general trend, we noticed some systematic differences between devices based on different chemical classes of materials used not only in the EML, but in adjacent transport layers as well. In principle, there was no reason to expect that all the diverse OLEDs would have the delayed emission due to a single mechanism. Therefore, we initially focused on blue fluorescent OLEDs, which to this day are still predominantly used in practical applications, primarily because of their relatively poor durability and colour of the phosphorescent OLEDs emitting blue light. These devices typically employed aromatic hydrocarbons such as anthracene derivatives as the main component (host) in the EML. It is noteworthy that, because of the bimolecular nature, the TTA process requires migration of triplets, which typically occurs via hopping between adjacent molecules. Therefore, the host material can be expected to play a central role in a putative TTA process. Considering the photophysical properties of anthracene in general and the prior observations of TTA in crystals of anthracene in particular, blue fluorescent OLEDs seem appropriate for an extensive study. Examination of various doped and undoped blue OLEDs showed the substantial delayed emission (initial intensity 3–32% of the normal electroluminescence) in almost all cases [25,26]. Representative excitations and decays of emission are depicted in figure 2 for several devices. Relatively high intensities of delayed emission (onsets identified by dashed lines in figure 2) allow direct acquisitions of normal and delayed electroluminescence in a single experiment without an

.........................................................

4. Ubiquitous nature of triplet–triplet annihilation in efficient fluorescent organic light-emitting diodes

6


From studies of electroluminescence in crystalline anthracene it is known that the TTA process may have a characteristic response to the magnetic field. It naturally follows that the similar magnetic field effects on OLED emission can be expected if a significant fraction of electroluminescence is due to TTA. Significant effects of the magnetic field in Alq3 -based devices were indeed found by Davis & Bussman [22], but the TTA involving molecular triplets appeared to be inconsistent with the experimental results. Although a later study [23] pointed out some uncertainties in interpreting the results, these uncertainties were deemed to be unlikely to affect the original conclusion about the negligible contribution of TTA in these experiments. The delayed emission in Alq3 -based OLEDs was studied in further detail by Popovic & Aziz [24], who pointed out important distinctions between delayed emissions due to TTA and those due to recombination of charge carriers retained in the device layers. They proposed application of the reverse bias to distinguish between these mechanisms and explained the intricate differences in experimental responses to reverse bias treatment by the presence of dopants and the hole-transporting material NPB in the emissive layer (EML). Similar to the results of Sinha & Monkman [21], their experimental technique precluded assessing the actual contribution of the TTA process to the overall electroluminescence. Nonetheless, usage of an optical chopper to prevent detector saturation suggests that the TTA-related emission is quite a minor component of electroluminescence. Overall, by the mid-2000s, there was no straightforward explanation of the higherthan-theoretical fluorescent OLED efficiencies (figure 1). The TTA mechanism, although not implausible in principle, has not been shown to be capable of producing efficiency gains anywhere close to what is required to explain the increasingly abundant experimental results. The alternative enhancement mechanisms were not ruled out in general and even less so for the growing number of examples of highly efficient fluorescent OLEDs.


7

1.0

normalized intensity

0.4

0.2

0 –50

–40

–30

–20

–10 time (ms)

0

10

20

30

Figure 2. Normalized time-resolved electroluminescence responses of representative fluorescent devices. Dashed lines identify the initial intensities of delayed components of electroluminescence. Devices are from the set depicted in fig. 11 of [26]. optical chopper. In all cases we examined, most of the electroluminescence is switched off within a fraction of a microsecond from the time the applied voltage is set below the turn-on voltage, typically 0 or −10 V. By contrast, the delayed emission decays much slower, on a microsecond or, in some cases, even orders of magnitude longer time scale. Because of the large difference in time scales of fast and slow components, it is straightforward to assess the initial intensity of the delayed emission, as exemplified by the arrow in figure 3. This figure also illustrates an important feature common to all the devices we examined: the delayed emission decay is largely unaffected by the turn-off bias as long as it is sufficiently low not to inject charge carriers. The slight difference between the decays in figure 3 is primarily due the RC time constant of the device affecting the time it takes for the normal electroluminescence to turn-off. The observed insensitivity of the decays to turn-off bias is a strong argument for TTA being the major source of the delayed emission. It is indeed to be expected that the electric field would have no effect on the bimolecular reaction involving two neutrally charged triplets. By contrast, the delayed recombination of the charge carriers potentially retained in the device after turn-off is likely to be affected by the electric field. Another noteworthy feature of the electroluminescence responses shown in figure 2 is that the electroluminescence rise also shows two distinct components: the fast one, occurring on the submicrosecond time scale, and the slow one, occurring on the microsecond time scale. This is qualitatively similar to the two components of the electroluminescence decay. Again, the fast component is largely due to the device RC constant, although injection and transport of carriers are likely to be significant contributors as well. Unlike with varying reverse bias when turning the device off, there was no clean method to control the turn-on process without affecting the slow component. Nonetheless, the slow component is sufficiently distinct to suggest the TTA process. In principle, assuming that there is a significant TTA contribution, the lifetime of triplets should play a similarly major role in both the rise (build-up of triplet to steady-state concentration) and the decay of the emission. In that case, it should be possible to relate the rise and decay processes quantitatively by variation of the excitation pulse width. In simple terms, if the pulse width is longer than the triplet lifetime, the triplet concentration is approaching its steady-state value and the initial intensity of the delayed emission is asymptotically approaching its maximum value. At the other extreme, triplet concentration is approximately proportional to (shorter) pulse

.........................................................

0.6


0.8


8

6

intensity (arb. units)

3 –10 V

0V

2 1

0

1

2

3 time (ms)

4

5

6

Figure 3. Initial part of electroluminescence decays (squares correspond to –10 V and crosses correspond to 0 V applied to switch the device off). The arrow identifies the initial intensity of the delayed component of electroluminescence. The device is described in [26] as A1. width and the initial intensity of the delayed emission is a quadratic function of the pulse width. Depending on whether the triplets decay predominantly by annihilation, kinetic models relating the pulse width and the initial intensity can be expressed as the two limiting case expressions [26]. Although the experimental data discussed below were generally described quite well by these kinetic models, it is noteworthy that the substantially more general treatments were provided by two research groups [27,28]. If the delayed emission is due to TTA, the triplet lifetimes in the ‘on’ state can be obtained from the kinetic models. A representative relationship between the triplet lifetimes and current densities is shown in figure 4. Although the lifetime appears nearly constant for low current densities, it is monotonically decreasing above approximately 1 mA cm−2 . This behaviour is consistent with an inherent decay (most likely, triplet diffusion followed by defect quenching) at low current densities and, at high current densities, with either a quenching of TTA-producing triplets by charge carriers or annihilation itself being the major channel of triplet consumption. By contrast, we were unable to describe satisfactorily the pulse width and current density effects in a framework of delayed recombination of charge carriers. It is instructive to relate the maximum value of the triplet state lifetime in the ‘on’ state lifetime (approx. 14 µs, in the case depicted in figure 4) to the approximately monoexponential decays observed in the ‘off’ state, e.g. after several microseconds after turning the device off. As shown in figure 5, the respective time constants are approximately 5–6 µs and are mostly independent of the pulse width. With respect to current density, there is a minor yet systematic increase in time constants with the decrease in current density. For example, in the case depicted in figure 5, lowering the current density from 5 to 0.64 mA cm−2 resulted in an approximately 25% increase in the time constant irrespective of pulse width. Because of the bimolecular nature of TTA, the observed time constant of the decay would be half of the triplet lifetime, which is therefore in the 10–12 µs range. Closely matched triplet lifetimes in the ‘on’ and ‘off’ state is another strong argument that the delayed emission is caused by the TTA in this exemplary case. As mentioned above, the observed effect of current density dependence on the triplet lifetime in the ‘on’ state can have two explanations in the framework of the TTA mechanism

.........................................................

4


5


16

9

lifetime (ms)

10 8 6 4 2 0 10–1

1

102

10 current density

(mA cm–2)

Figure 4. Triplet state lifetime obtained from kinetics models in the ‘on’ state as a function of current density. The device is described in [26] as A1.

intensity (arb. units)

10

1

10–1

10–2 0

0.000002

0.000004

0.000006

0.000008 time (s)

0.000010

0.000012

0.000014

Figure 5. Electroluminescence decays obtained with different pulse widths (top to bottom: 100, 8, 5, 4 µs) for the device described in [26] as A1. Dashed lines are eye guides to point out the monoexponential character of the decays. of delayed electroluminescence. In devices with a relatively low contribution of TTA to the total electroluminescence, it is reasonable to assume that the annihilation is never a dominant mechanism of triplet decay and, therefore, the change in triplet lifetime must be due to a current density-related increase in concentrations of charge carriers, which are effective quenchers of triplet states. In fact, hole transport material- and electron transport material-derived charge carriers were directly shown to quench TTA-inducing triplets in several different OLED devices [25,26]. On the other hand, in the devices where the TTA contribution is high, the consumption of

.........................................................

12


14


In principle, the collision of two uncorrelated triplets (T1 ) in the absence of a magnetic field may result in nine equally probable statistical outcomes, one of which is singlet (S), three are triplets (T) and five are quintets (Q) [34], 1 5 1 T1 + T1 → S + T + Q. 9 3 9

.........................................................

5. Limits of triplet–triplet annihilation-related efficiency enhancement

10


triplets may be second order in triplet concentration and the (pseudo)first-order time constant— triplet lifetime—is naturally changing with current density. Another general observation suggesting the TTA mechanism is the distinct shape of the dependence of initial intensity of delayed emission on current density. As expected, the intensity scales approximately quadratically with current density below a certain point, and approximately linearly above it. The low current density regime is easily explained by the expected first-order decay (radiative and non-radiative) of excited states, which results in the triplet concentration being a linear function of the generation rate (or current density), and the TTA intensity being, respectively, a quadratic function. Obviously, such a quadratic dependence may not persist ad infinitum. As current density increases, the TTA contribution may be expected to turn into an approximately linear function of current density when the TTA becomes the dominant triplet decay process. Interestingly, the same behaviour is sometimes observed even when the TTA contribution is quite low and cannot be expected to dominate triplet consumption. For an example of such behaviour, see fig. 4 in [25]. It is likely that, in such cases, charge carrier quenching is the limiting process that is responsible for approximately linear dependence at high current densities. In view of the evidence supporting a major role of TTA in delayed emission, the electroluminescence of efficient fluorescent devices is expected to generally consist of two components: the TTA-related one, produced by excited singlets formed from annihilating triplets, and the one due to excited singlets produced directly by regular recombination. The intensity of the former component can be determined from the leading part of the delayed emission and the intensity of latter by subtracting the TTA component from the total intensity. As mentioned above, we observed a general, albeit noisy, trend between the efficiencies of fluorescent devices and their delayed emission intensities. Considering the low end of the trend, it is not surprising that the inefficient fluorescent devices typically have a low TTA contribution too: most shortcomings in device design leading to decreased production or excessive quenching of singlet excited states can be plausibly expected to affect triplet states too. At the other extreme of the trend, we observed that the highest TTA contributions, approximately half of the total emission, were observed in red fluorescent devices with a exceptionally high efficiencies [29]. The efficiency–TTA trend suggests that the efficiency differences may be to a large degree related to the extent of the TTA. By way of illustration, in a set of architecturally similar devices where only the electron transport layer is modified, we observed that there is indeed a clear correlation between efficiency (EQE) and the TTA-related EQE component [26]. As depicted in figure 6, the slope of the trendline is nearly unity, which suggests that not only is the TTA a significant contributor to efficiency but also that it dominates variations in efficiency in this case. Within a few years following our findings that TTA is a substantial efficiency contributor [25,26,30], several R&D teams from the major organizations developing fluorescent OLEDs reported similar results in their proprietary, state-of-the-art devices [3,31–33]. Also consistent with our observations, the importance of the layers immediately adjacent to the EML was emphasized. Interestingly, despite the large differences in device architectures and the chemical nature of the EML materials, the polymer-based OLEDs benefited from TTA to a similar degree [32,33]. Overall, we conclude that the TTA is truly ubiquitous in efficient fluorescent OLEDs. One can speculate that the empirical optimization process, which has efficiency as one of the major drivers, almost unavoidably leads to the development of devices benefiting from substantial TTA-related efficiency enhancement.


1

0 6.0

6.5

7.0

7.5 EQE (%)

8.0

8.5

9.0

Figure 6. TTA contribution as a function of EQE. The solid line identifies the correlation between the TTA contribution and EQE. Devices are from the set depicted in fig. 11 of [26].

In contrast to the recombination of charge carriers, where coulombic attraction makes subsequent separation unlikely, a substantial fraction of triplet encounter pairs can separate, thereby regenerating free triplets. In fact, the high endothermicity of quintet state formation prevents annihilation of triplet pairs that would otherwise lead to molecular quintets for many organic molecules, including anthracene and tetracene derivatives [29,35]. For example, in the case of rubrene, the density functional calculations show that formation of the lowest molecular quintet state from two triplets is deficient in energy by more than 2 eV [29]. This prevents annihilation of such pairs and changes the overall spin statistics as follows: 3 1 T1 + T1 → S + T. 4 4 Considering that the annihilation events resulting in triplet formation will generally consume only one triplet, it is easy to see that the spin statics yield one singlet state per five annihilated triplets. In the context of fluorescent OLEDs where recombination produces 25% of singlets and 75% of triplets, the TTA process therefore may add 75/5 = 15% of singlets and the total yield of singlets may become 40%. As pointed out in the Introduction, it is problematic to measure yields of photons or emissive excited states accurately. By contrast, measuring fractions of excited singlets produced by TTA is quite straightforward. Neglecting various loss processes, which might quench singlets and triplets to a different degree, the maximum TTA contribution is expected to be 15/40 = 37.5% of total electroluminescence based on spin statistics. This value is in good agreement with the experimental results obtained in blue fluorescent devices: the highest reported TTA contributions are approximately 35% [31,36], or, more typically, are only slightly above 30%. Because the TTA-produced singlets are subject to the same quenching mechanisms as singlets generated directly in recombination and furthermore affected by quenching of triplets, it is indeed reasonable to expect the experimental TTA contribution to be lower than the predicted maximum. Nonetheless, the differences in spatial distribution of TTA-produced and recombination-produced may, in accidentally or intentionally constructed special cases, result in outcoupling-related suppression of direct-recombination-produced emission [26]. As a result, the experimental TTA contribution may even substantially exceed the 37.5% predicted maximum, but the overall efficiency would be necessarily decreased in those cases.

.........................................................

2


TTA-related EQE component (%)

11


T1 + T1

TTS

k–1 kT

TTT

k–1

singlet channel

S1 + S 0 T1 + T 1

triplet channel (different energetics, kT π kS)

Tn + S 0 T1 + T 1

dS1/dTn = 1/3 * kS/kT* (kT + k–1)/(kS + k–1)

branching ratio

if most triplet pairs separate:

dS1/dTn ~ 1/3 * kS/kT

if most triplet pairs annihilate:

dS1/dTn ~ 1/3 (independent of the kS/kT ratio)

Scheme 1. Kinetic depiction of the relationship between the singlet and triplet channels of TTA.

Interestingly, the exceptionally efficient (nearly 12% EQE) red fluorescent OLEDs reported separately by the research groups from Kodak [10] and Novaled [3] appear to exceed the 37.5% limit [29]. Similar to our observations for blue devices, the delayed emission showed all the clear signatures of being TTA-generated, and yet its intensity was approximately half of the total emission in those devices. On the one hand, this TTA contribution, taken at its face value, explains the high efficiency: without the TTA-related component, the EQE is only approximately 6%, which is quite reasonable for a fluorescent device without enhanced outcoupling. On the other hand, the combination of high efficiency and high TTA contribution is plainly irreconcilable with the 37.5% limit derived from spin statistics. However, the spin-statistical limit only applies if an encounter pair with either triplet or singlet character forms irreversibly and yields a final molecular excited state without change in singlet : triplet probabilities. With such a constraint of irreversibility, the rates of conversion of the encounter pair into triplet and singlet molecular excited states, even if different, would not affect the 3 : 1 ratio between triplet and singlet products. Considering an absence of coulombic attractions between two triplets, there is a plausible alternative that an encounter pair forms reversibly and may regenerate two original triplets even when the annihilation is energetically allowed. In that case, the difference in rate constants of the formation of final singlet or triplet states would change the ratio as according to Scheme 1. Such reversibility was experimentally demonstrated by Merrifield [34] in an example of anthracene where 4% of the encounters of two triplets led to fusion via the singlet channel, 13% via the triplet channel and 83% scatter without reaction. Furthermore, the large differences in the energetics of singlet- and tripletyielding annihilation reactions can be expected to be reflected in the respective rate constants. This appears to be the case for rubrene and structurally related molecules, which are used as hosts in the red fluorescent devices with exceptional efficiencies. The very close match between twice the triplet state energy and the first excited state energy is likely to be the fundamental reason for kT < kS and, consequently, for the large deviation of the spin-statistical ratio in favour of singlets. An independent confirmation of the substantial violation of spin statistics was obtained in a study of magnetic field effects. Because the magnetic field only affects the singlet channel [34], it can be used to examine the competition between the two annihilation channels and determine their relative contributions to TTA. Analysis of time-resolved electroluminescence with and without the magnetic field yielded a 1.3 branching ratio, which is substantially higher than the value of 0.33 expected for a purely statistical annihilation process [29]. Interestingly, the branching ratio obtained in this study of magnetic field effects was in excellent agreement with the observed fraction of TTA-related emission in total electroluminescence.

12 .........................................................

1/3k1

kS


T1 + T1

1/9k1


13 .........................................................


It may be concluded that the TTA process in practical fluorescent devices is, at least in principle, capable of more than doubling singlet yield from the classical estimate of 25% to almost (25 + 75/2) = 62.5%. Furthermore, enhancing the singlet channel of TTA opens up the possibility of creating fluorescent OLEDs approaching 100% power efficiency (neglecting outcoupling losses): it has been demonstrated that the rubrene/PTCDI (N, N -ditridecylperylene-3,4,9, 10-tetracarboxylic diimide) heterojunction device produces yellow electroluminescence with low efficiency (EQE approx. 0.1%) and extraordinarily low turn-on voltage (approx. 1 V), which is slightly below half of the rubrene excited singlet state energy [37]. Considering the large energy deficit (more than 1 eV), the recombination is obviously unable to produce an emissive state of rubrene directly in such a device, which led the authors to postulate a rather doubtful Augerlike two-step injection mechanism to explain the nature of electroluminescence. However, the observed electroluminescence can be readily understood without invoking any new mechanisms if one notices that the recombination leading to the triplet state of rubrene is not forbidden by energy, and, furthermore, the rubrene triplets readily undergo TTA to generate the emissive singlet state. In fact, by using a similar heterojunction in doped OLEDs resembling highly efficient OLEDs [29], we have been able to construct red-emitting devices operating relatively efficiently at similarly low voltages and demonstrate that their electroluminescence is dominated by the delayed component that shows all typical tell-tale signs of TTA (T. D. Pawlik, D. Y. Kondakov 2009, unpublished data). Even though the TTA in a device of this type cannot produce more than 50% of the excited singlet yield, these singlets may be produced at approximately half the voltage, thereby demonstrating that it may indeed be possible to construct a fluorescent OLED comparable to phosphorescent OLEDs in terms of power efficiency. Unfortunately, realizing devices with the substantially enhanced TTA requires molecules specially designed to be capable of altering the spin-statistical ratio in TTA. An example of such a molecule is rubrene, which is compatible with the red fluorescent devices as an EML host. No functional analogues capable of performing as blue or green OLED hosts have been reported so far. Therefore, the TTA-related efficiency gains for the devices that are most important for practical use—fluorescent blue OLEDs—are substantially less impressive, with the TTA contribution somewhat below the statistical 37.5% and the total singlet yield below 40%. In the majority of practical devices, the TTA contributions only slightly exceed 30% in the efficient fluorescent blue OLEDs. It is perhaps an even more typical scenario that the TTA contribution is between 20% and 30%. As mentioned above, it is likely that the increases in the TTA extent were primarily driven by extensive empirical optimizations of OLED efficiencies conducted by various research groups in parallel. If so, why did not such optimization drive the TTA even higher towards the spin-statistical limit? One noteworthy reason is that the empirical optimization of OLEDs is not driven solely by the efficiency. The durability or, more precisely, operational lifetime is another major characteristic of OLEDs that was subject to more than two decades of optimization. Interestingly, the studies of TTA extent in small molecule and polymeric devices revealed that the TTA-producing triplets are strongly affected by operation-related device degradation [24,32,33]. As a result, in addition to various loss mechanisms affecting excited singlets irrespective of how they were produced, the TTA-related component of electroluminescence is furthermore decreased by loss processes specific to triplet states. As pointed out by Roberts et al. [32,33], this may cause the TTA-related component of electroluminescence to decrease faster than the component due to singlets directly produced by recombination and, therefore, result in a luminance–time curve showing rapid decline until most of the TTA is extinguished. This behaviour is undesirable because it has disproportionately large negative effects on the more important lifetime characteristics describing early luminance decay: the time to reach 97% or 95% of initial luminance, in particular. Consequently, an intentional elimination or reduction of TTA, for example by adding triplet quenchers, can be used to increase the fluorescent device’s lifetime, but at the expense of reduced efficiency [32,33]. Similar trade-offs are probably achieved in the empirical process of OLED device development, which may explain the lower TTA contributions typically found in the majority of practical OLEDs.


6. Concluding remarks

1. Tang CW, VanSlyke SA. 1987 Organic electroluminescent diodes. Appl. Phys. Lett. 51, 913–915. (doi:10.1063/1.98799) 2. Böhm E, Pflumm C, Voges F, Flämmich M, Heil H, Büsing A, Parham A, Fortte R, Mujica T. 2009 Novel transport materials for high performance OLEDs. In Proc. 16th Int. Display Workshops Conf., Miyazaki, Japan, 9–11 December 2009. Campbell, CA: Society for Information Display. 3. Rothe C, Werner A, Fadhel O, Limmert M, Canzler TW, Birnstock J. 2009 Improved efficiency and lifetime by tailoring the charge carrier supply in OLEDs. SID Int. Symp. Digest Tech. Papers 40, 495–498. (doi:10.1889/1.3256824) 4. Kondakov DY, Young RH. 2010 Variable sensitivity of organic light-emitting diodes to operation-induced chemical degradation: nature of the antagonistic relationship between lifetime and efficiency. J. Appl. Phys. 108, 074513. (doi:10.1063/1.3483251) 5. Burroughes JH, Bradley DDC, Brown AR, Marks RN, Mackay K, Friend RH, Burns PL, Holmes AB. 1990 Light-emitting diodes based on conjugated polymers. Nature 347, 539–541. (doi:10.1038/347539a0) 6. Tsutsui T. 1997 Progress in electroluminescent devices using molecular thin films. Mater. Res. Bull. 22, 39–45. 7. Kido J, Iizumi Y. 1998 Fabrication of highly efficient organic electroluminescent devices. Appl. Phys. Lett. 73, 2721–2723. (doi:10.1063/1.122570) 8. Liao CH, Lee MT, Tsai CH, Chen CH. 2005 Highly efficient blue organic light-emitting devices incorporating a composite hole transport layer. Appl. Phys. Lett. 86, 203507/1–203507/3. 9. Okumoto K, Kanno H, Hamaa Y, Takahashi H, Shibata K. 2006 Green fluorescent organic light-emitting device with external quantum efficiency of nearly 10%. Appl. Phys. Lett. 89, 063504.

.........................................................

References


Towards the end of the 2000s, the essential role of TTA in realizing highly efficient fluorescent devices was widely recognized. The fundamental features of OLED efficiency enhancement via TTA—molecular structure-dependent contribution maxima and mechanisms reducing them in practical devices as a function of current density [25,26,28,38,39] as well as the typical antagonistic connection between TTA and OLED lifetime [32,33]—came to be understood over the course of the next few years. Afterwards, however, there was much less progress with respect to the all-important quantitative details of the TTA mechanism. For example, it is well understood that, for a given OLED, there is usually some current density below which the TTA-related component of electroluminescence declines rapidly (approximately quadratically with respect to current density), thereby contributing to overall efficiency change with current density. Yet there is virtually no understanding of how the device architecture and the properties of OLED materials affect at what current density this change occurs. Even the most basic property—maximum TTA contribution—generally cannot be predicted from the device architecture and properties of OLED materials. There is a striking contrast between the level of activity in developing materials for the singlet fission process for organic photovoltaics (for a recent review, see [40]) and the fundamentally related process of TTA producing predominantly excited singlet states. It should be emphasized that, to this day and despite the decades of work on improving blue phosphorescent OLEDs as well as the recent advent of the TADF OLEDs, the majority of practical display and lighting OLEDs still rely on TTA for creating blue light efficiently. Considering such practical importance of fluorescent blue OLEDs, the apparent lack of mechanistic understanding and ability to control the quantitative aspects of TTA clearly calls for further enhanced effort in this area. The development of blue-OLED-compatible materials capable of substantially exceeding the spin-statistical limit in TTA, the elimination of the antagonistic relationship between TTA-related gains in efficiency and losses in lifetime, and the design of devices with an extended range of current densities producing near-maximum TTA electroluminescence are currently the areas where improvements would be most beneficial.

14


15 .........................................................


10. Spindler JP, Begley WJ, Hatwar TK, Kondakov DY. 2009 High-efficiency fluorescent red- and yellow-emitting OLED devices. SID Int. Symp. Digest Tech. Papers 40, 420–423. (doi:10.1889/ 1.3256804) 11. Deaton JC et al. 2010 E-type delayed fluorescence of a phosphine-supported Cu2 (μ-NAr2 )2 diamond core: harvesting singlet and triplet excitons in OLEDs. J. Am. Chem. Soc. 132, 9499– 9508. (doi:10.1021/ja1004575) 12. Uoyama H, Goushi K, Shizu K, Nomura H, Adachi C. 2012 Highly efficient organic lightemitting diodes from delayed fluorescence. Nature 492, 234–238. (doi:10.1038/nature11687) 13. Segal M, Baldo MA, Holmes RJ, Forrest SR, Soos ZG. 2003 Excitonic singlet–triplet ratios in molecular and polymeric organic materials. Phys. Rev. 68, 075211. (doi:10.1103/ PhysRevB.68.075211) 14. Cao Y, Parker ID, Zhang C, Heeger AJ. 1999 Improved quantum efficiency for electroluminescence in semiconducting polymers. Nature 397, 414–417. (doi:10.1038/17087) 15. Wohlgenannt M, Tandon K, Mazumdar S, Ramasesha S, Vardeny ZV. 2001 Formation crosssections of singlet and triplet excitons in π -conjugated polymers. Nature 409, 494–497. (doi:10.1038/35054025) 16. Tandon K, Ramasesha S, Mazumdar S. 2003 Electron correlation effects in electronhole recombination in organic lightemitting diodes. Phys. Rev. B 67, 045109. (doi:10. L1103/PhysRevB.67.045109) 17. Monkman AP. 2013 Singlet generation from triplet excitons in fluorescent organic lightemitting diodes. ISRN Mater. Sci. 2013, 670130. (doi:10.1155/2013/670130) 18. Segal M, Singh M, Rivoire K, Difley S, Van Voorhis T, Baldo MA. 2007 Extrafluorescent electroluminescence in organic light-emitting devices. Nat. Mater. 6, 374–378. (doi:10. 1038/nmat1885) 19. Hochstrasser RM. 1962 Luminescence of organic molecular crystals. Rev. Mod. Phys. 34, 531– 550. (doi:10.1103/RevModPhys.34.531) 20. Ganzorig C, Fujihira M. 2002 A possible mechanism for enhanced electrofluorescence emission through triplet–triplet annihilation in organic electroluminescent devices. Appl. Phys. Lett. 81, 3137–3139. (doi:10.1063/1.1515129) 21. Sinha S, Monkman AP. 2003 Delayed electroluminescence via triplet–triplet annihilation in light emitting diodes based on poly[2-methoxy-5-(2-ethyl-hexyloxy)-1,4-phenylene vinylene. Appl. Phys. Lett. 82, 4651–4653. (doi:10.1063/1.1588733) 22. Davis AH, Bussman K. 2004 Large magnetic field effects in organic light emitting diodes based on tris(8-hydroxyquinoline aluminum) (Alq3 )/N, N -di(naphthalen-1-yl)N, N‘diphenyl-benzidine (NPB) bilayers. J. Vac. Sci. Tech. A 22, 1885–1891. (doi:10.1116/ 1.1759347) 23. Wilkinson J, Davis AH, Bussmann K, Long JP. 2005 Evidence for charge-carrier mediated magnetic-field modulation of electroluminescence in organic light-emitting diodes. Appl. Phys. Lett. 86, 111109. (doi:10.1063/1.1883322) 24. Popovic ZD, Aziz H. 2005 Delayed electroluminescence in small-molecule-based organic light-emitting diodes: evidence for triplet–triplet annihilation and recombinationcenter-mediated light-generation mechanism. J. Appl. Phys. 98, 013510. (doi:10.1063/1. 1937472) 25. Kondakov DY. 2007 Characterization of triplet–triplet annihilation in organic lightemitting diodes based on anthracene derivatives. J. Appl. Phys. 102, 114504. (doi:10.1063/ 1.2818362) 26. Kondakov DY. 2009 Role of triplet–triplet annihilation in highly efficient fluorescent devices. J. Soc. Inf. Disp. 17, 137–144. (doi:10.1889/JSID17.2.137) 27. Rothe C, King SM, Monkman AP. 2006 Direct measurement of the singlet generation yield in polymer light-emitting diodes. Phys. Rev. Lett. 97, 076602. (doi:10.1103/PhysRevLett. 97.076602) 28. Wallikewitz BH, Kabra D, Gelinas S, Friend RH. 2012 Triplet dynamics in fluorescent polymer light-emitting diodes. Phys. Rev. B 85, 045209. (doi:10.1103/PhysRevB.85. 045209) 29. Kondakov DY, Pawlik TD, Hatwar TK, Spindler JP. 2009 Triplet annihilation exceeding spin statistical limit in highly efficient fluorescent organic light-emitting diodes. J. Appl. Phys. 106, 124510. (doi:10.1063/1.3273407) 30. Kondakov DY. 2008 Role of triplet–triplet annihilation in highly efficient fluorescent devices. SID Symp. Digest Tech. Papers 39, 617–620. (doi:10.1889/1.3069738)


16 .........................................................


31. Kawamura M, Kawamura Y, Mizuki Y, Funahashi M, Kuma H, Hosokawa C. 2010 Highly efficient fluorescent blue OLEDs with efficiency-enhancement layer. SID Symp. Digest Tech. Papers 41, 560–563. (doi:10.1889/1.3500528) 32. Roberts M, King S, Cass M, Pintani M, Coward C, Akino N, Nakajima H, Anryu M. 2011 Excited state interactions in P-OLEDs: implications for efficiency and lifetime. SID Symp. Digest Tech. Papers 42, 820–821. (doi:10.1889/1.3621456) 33. King SM, Cass M, Pintani M, Coward C, Dias FB, Monkman AP, Roberts M. 2011 The contribution of triplet–triplet annihilation to the lifetime and efficiency of fluorescent polymer organic light emitting diodes. J. Appl. Phys. 109, 074502. (doi:10.1063/1.3561430) 34. Merrifield RE. 1968 Theory of magnetic field effects on the mutual annihilation of triplet excitons. J. Chem. Phys. 48, 4318–4323. (doi:10.1063/1.1669777) 35. Dick B, Nickel B. 1983 Accessibility of the lowest quintet state of organic molecules through triplet–triplet annihilation; an indo ci study. Chem. Phys. 78, 1–16. (doi:10.1016/03010104(83)87001-3) 36. Yamada T. 2013 Recent progress on polymer OLED materials. In Proc. In-depth Technology Insight Conf., Seoul, Korea, 2013. 37. Pandey AK, Nunzia JM. 2007 Upconversion injection in rubrene/perylene-diimideheterostructure electroluminescent diodes. Appl. Phys. Lett. 90, 263508. (doi:10.1063/ 1.2752540) 38. Zhang Y, Forrest SR. 2012 Triplets contribute to both an increase and loss in fluorescent yield in organic light emitting diodes. Phys. Rev. Lett. 108, 2 674 041–2 674 045. 39. Luo Y, Aziz H. 2010 Probing triplet–triplet annihilation zone and determining triplet exciton diffusion length by using delayed electroluminescence. J. Appl. Phys. 107, 094510. (doi:10.1063/1.3410678) 40. Smith MB, Michl J. 2010 Singlet fission. Chem. Rev. 110, 6891–6936. (doi:10.1021/cr1002613)

Meningococcal vaccines: Current state and future outlook.

Highly Efficient White Organic Light-Emitting Diodes with Ultrathin Emissive Layers and a Spacer-Free Structure.

Highly efficient and stable electron injection layer for inverted organic light-emitting diodes.

Highly efficient and stable organic light-emitting diodes with a greatly reduced amount of phosphorescent emitter.

Highly efficient exciplex organic light-emitting diodes incorporating a heptazine derivative as an electron acceptor.

Highly efficient blue organic light-emitting diodes using quantum well-like multiple emissive layer structure.

An efficient nano-composite layer for highly transparent organic light emitting diodes.

Low-dose antibiotics: current status and outlook for the future.

Stem cells for osteodegenerative diseases: current studies and future outlook.

High-efficiency organic light-emitting diodes with fluorescent emitters.

Treatment of Cannabis Use Disorder: Current Science and Future Outlook.

Highly efficient organic light-emitting diodes with phosphorescent emitters having high quantum yield and horizontal orientation of transition dipole moments.

An easy route to red emitting homoleptic IrIII complex for highly efficient solution-processed phosphorescent organic light-emitting diodes.

Highly efficient, solution processed electrofluorescent small molecule white organic light-emitting diodes with a hybrid electron injection layer.

Neuroimaging trends and future outlook.

Highly efficient organic light emitting diodes formed by solution processed red emitters with evaporated blue common layer structure.

Advances in the Classification and Treatment of Mastocytosis: Current Status and Outlook toward the Future.

An optical and electrical study of full thermally activated delayed fluorescent white organic light-emitting diodes.

Host Engineering for High Quantum Efficiency Blue and White Fluorescent Organic Light-Emitting Diodes.

Heart transplantation: current outlook.

Treating Alzheimer's disease with monoclonal antibodies: current status and outlook for the future.

Transient elastography in Canada: current state and future directions.

Activity-based proteomics probes for carbohydrate-processing enzymes: current trends and future outlook.

Current state and future prospects for psychosomatic medicine in Japan.