
==== Front
Research (Wash D C)
Research (Wash D C)
RESEARCH
Research
2639-5274
AAAS

10.34133/research.0009
0009
Research Article
White Fluorescent Organic Light-Emitting Diodes with 100% Power Conversion
Ding Dongxue †
Wang Zicheng †
Duan Chunbo
Han Chunmiao
Zhang Jing
Chen Shuo
Wei Ying *
Xu Hui *
Key Laboratory of Functional Inorganic Material Chemistry, Ministry of Education, School of Chemistry and Materials, Heilongjiang University, Harbin, Heilongjiang, China.
* Address correspondence to: ywei@hlju.edu.cn (Y.W.); hxu@hlju.edu.cn (H.X.)
† These authors contributed equally to this work.

19 12 2022
2022
2022 000918 8 2022
21 10 2022
Copyright © 2022 Dongxue Ding et al.
2022
Dongxue Ding et al.
https://creativecommons.org/licenses/by/4.0/ Exclusive licensee Science and Technology Review Publishing House. No claim to original U.S. Government Works. Distributed under a Creative Commons Attribution License 4.0 (CC BY 4.0).

Energy-efficient lighting sources are desired to provide another solution of carbon emission reduction. White organic light-emitting diodes are promising, because of theoretical internal quantum efficiencies for 100% electric-to-light conversion. However, pure organic fluorescent materials still face a challenge in harvesting triplet excitons for radiation. Herein, we report a white fluorescent organic light-emitting diode having an external quantum efficiency of 30.7% and a power efficiency of 120.2 lm W−1. In the single emissive layers, we use blue thermally activated delayed fluorescent emitters to sensitize a yellow fluorescent emitter. Transient photoluminescence and electroluminescence analyses suggest that a blue thermally activated delayed fluorescent molecule with ~100% reverse intersystem crossing efficiency and negligible triplet nonradiative rate constant completely converts triplet to singlet, suppressing triplet quenching by a yellow fluorescent emitter and ensuring 100% power conversion.
==== Body
pmcIntroduction

White organic light-emitting diodes (OLEDs) based on pure organic materials are promising as highly efficient large-area lighting sources, because of the merits of low cost, environmental sustainability, and large-scale production [1,2]. However, according to spin statistics of electrogenerated excitons, organic fluorophores can use only 25% singlet excitons [3]. In contrast, phosphorescent organometallic complexes of noble metals can utilize 100% excitons of both singlet and triplet, through intersystem crossing (ISC) [4]. Recently, thermally activated delayed fluorescence (TADF) has emerged as a competitive alternative for harvesting 100% excitons for electroluminescence (EL) [5,6]. Most of TADF molecules are pure organic, featuring donor–acceptor structures and near-zero singlet–triplet energy gaps [7,8]. In this case, reverse intersystem crossing (RISC) can convert the nonradiative triplet of the TADF molecule to a radiative singlet, leading to 100% exciton harvesting [9–11]. Monochromic TADF diodes have realized internal quantum efficiency (IQE, ηIQE) and external quantum efficiency (EQE, ηEQE) respectively approaching 100% and 30%, which are already comparable to the values of state-of-the-art phosphorescent OLEDs [12–16]. Despite the high singlet radiative rate and feasible RISC, the triplet is still dominant in the singlet–triplet equilibrium, because of the kinetic advantage of exothermic ISC to endothermic RISC [17]. Thus, TADF emitters suffer from both triplet–triplet annihilation and singlet–triplet quenching [18], inducing more marked efficiency roll-offs of TADF diodes than phosphorescent counterparts [19–21]. Due to the collisional nature of quenching effects, TADF molecules are highly sensitive to intermolecular interactions, especially for blue emitters [22–25]. Although there are efficient TADF-phosphorescence hybrid [26–28] and full-TADF white OLEDs [29–31] reported, the interactions between blue and green/yellow/red TADF emitters are quite complicated [32], requiring accurate optimization of molecular configurations and device structures [2,33,34].

Their long-lived feature makes triplets inherently easy to be quenched by intermolecular interactions. Therefore, enhancing triplet-to-singlet conversion is a fundamental way to restrain quenching-induced efficiency reduction and roll-off [35–37]. It is demonstrated that TADF molecules can be used as triplet sensitizers of traditional fluorescent (FL) emitters [38,39]. The so-called “hyperfluorescence” diodes with emissive layers containing FL and TADF molecules can also achieve ηEQE more than 20%. In these devices, singlet excitons converted from the first triplet excited state (T1) of TADF sensitizers can be subsequently captured and utilized by FL emitters, through Förster resonance energy transfer (FRET). This process not only makes singlet–triplet equilibrium become more favorable to singlet formation but also transforms relatively labile charge transfer excitons of TADF molecules into stable Frenkel excitons of FL emitters. It is beneficial to comprehensively mitigate nonradiative deactivation of excitons [3]. However, triplet energy transfer from TADF sensitizers to the dark T1 state of FL emitters should be thoroughly avoided. Dexter energy transfer (DET) of triplets is based on short-distance charge exchange (within 1 nm). Thus, low doping concentration of FL emitters is believed to be necessary and sufficient to maintain a long enough average distance between TADF and FL molecules and thereby prevent triplet DET (Fig. 1A). However, at present, hyperfluorescence OLEDs hardly achieved ηEQE reaching the state-of-the-art value of TADF diodes (30%). It means that a part of excitons was still wasted during the energy transfer process [40,41].

Fig. 1. Design strategy of hyperfluorescence white-emitting systems. (A) Spatial features of triplet energy transfer and diffusion in a white hyperfluorescence system consisting of a yellow FL emitter, a blue TADF sensitizer, and a host matrix. (B) Key transitions and energy transfer processes between the blue TADF sensitizer and the yellow FL emitter. (C) Chemical structures of 9-(4-(4,6-Bis(4-(diphenylphosphoryl)phenyl)-1,3,5-triazin-2-yl)phenyl)-3,6-di-tert-butyl-carbazole (ptBCzPO2TPTZ), 4,5-bis(carbazol-9-yl)-1,2-dicyanobenzene (2CzPN), and Bis[4-(9,9-dimethyl-9,10-dihydroacridine)phenyl]sulfone (DMAC-DPS), and RISC efficiencies (ϕRISC) and triplet nonradiative rate constants (knrT) of their 40% doped 4,6-bis(diphenylphosphoryl)dibenzofuran (DBFDPO) films (vacuum evaporated, 100 nm). (D) Electronic absorption spectra of ptBCzPO2TPTZ, 2CzPN, DMAC-DPS, and 2,8-di-tert-butyl-5,11-bis(4-tert-butylphenyl)-6,12-diphenyltetracene (TBRb) in dichloromethane (10-6 mol l−1) and photoluminescence (PL) spectra of ptBCzPO2TPTZ, 2CzPN, DMAC-DPS, and TBRb doped in DBFDPO films (40%, vacuum evaporated, 100 nm). Inset shows the chemical structure of TBRb.

Charge transfer excitons of TADF molecules are much easier to long-range delocalize, compared to Frenkel excitons. Besides DET, another feature of triplet excitons must be noted: triplet diffusion distance (≥10 nm) is at least one order of magnitude larger than singlet diffusion distance (≤1 nm) [42] (Fig. 1A). In this case, triplet diffusion actually provides another channel for triplet capture by FL emitters. Triplet DET and triplet diffusion are respectively dominant at relatively high and low doping concentration. The energy level relationship between TADF sensitizers and FL emitters shows that it is difficult to completely eradicate triplet migration between their T1 states, since effective FRET should be based on a suitable concentration of FL emitters (Fig. 1B). Therefore, an “ideal” situation is triplet-involved processes confined on TADF molecules, which requires the following: (a) the TADF sensitizer completely converts triplet to singlet, and (b) before conversion, triplet nonradiation is effectively suppressed. As a proof of concept, in this contribution, we demonstrate a white hyperfluorescence system featuring triplet-free exciton allocation. Three blue TADF emitters named ptBCzPO2TPTZ, 2CzPN, and DMAC-DPS are respectively used as blue-emitting sensitizers to fabricate hyperfluorescence white OLEDs with a conventional yellow FL emitter TBRb (Fig. 1C). These 3 molecules have similar molecular polarities, thus excluding the influence of dipole–dipole interactions between TADF and FL emitters. In the 4,6-bis(diphenylphosphoryl)dibenzofuran (DBFDPO) matrix with a doping concentration of 40% for weight percentage, the RISC efficiency (ϕRISC) and triplet nonradiative rate constant (knrT) of ptBCzPO2TPTZ are 97% and 2.15 × 103 s−1, respectively. In contrast, ϕRISC values of 2CzPN and DMAC-DPS (~80%) are much lower, while the knrT value of the 2CzPN-based film (2.24 × 103 s−1) is equal to that of ptBCzPO2TPTZ, but is only one-fifth of that of the DMAC-DPS-based film (1.26 × 104 s−1). The orthogonal correlation of ϕRISC and knrT values for these molecules establishes the basis for figuring out exciton utilization and quenching processes and the relative structure–property relationships. Blue emissions from ptBCzPO2TPTZ, 2CzPN, and DMAC-DPS are complementary to the yellow emission of TBRb, and simultaneously overlap with the π→π* absorption band of TBRb in a large range of 400 to 550 nm, leading to efficient FRET (Fig. 1D). Compared to 2CzPN and DMAC-DPS, ptBCzPO2TPTZ with ~100% ϕRISC and negligible knrT effectively mitigates triplet quenching by TBRb. As a result, warm-white OLED of DBFDPO:40% ptBCzPO2TPTZ:0.1% TBRb achieved a state-of-the-art ηEQE of up to 30.7%, corresponding to 100% ηIQE, and a record-high power efficiency (ηPE) of 120.2 lm W−1.

Results

Steady-state photophysical properties

The photoluminescence (PL) quantum yield (ηPL) of the DBFDPO:40% ptBCzPO2TPTZ film reaches 94%, which is higher than ~70% of 2CzPN- and DMAC-DPS-doped films (Fig. 2A to C). More importantly, for DBFDPO:x% ptBCzPO2TPTZ:y% TBRb dually doped films, at x = 40, a slight codoping of 0.1% TBRb largely increases ηPL to 99%, reflecting nearly unitary radiative efficiency and completely suppressed nonradiative transitions. However, further increasing TBRb concentration gradually decreases ηPL. Similarly, when x ≤ 40, the ηPL of DBFDPO:x% ptBCzPO2TPTZ:0.1% TBRb films gradually increases and reaches the maximum at x = 40 (Fig. S1). Further increasing x also induces ηPL decrease. The ηPL inflexion is a result of a delicate balance between exciton radiation and quenching. In contrast, for DBFDPO:40% 2CzPN:y% TBRb films, ηPL decreases to 47% even at y = 0.1, and further halved to ~20% at y ≥ 0.2. The quenching effect of TBRb is the worst for DBFDPO:40% DMAC-DPS:y% TBRb films, whose ηPL is as low as ~10% when y ≥ 0.1. Notably, ptBCzPO2TPTZ-based films reveal gradually reduced ηPL inversely proportional to x and y, reflecting a linear dependence of triplet quenching on ptBCzPO2TPTZ–TBRb distance. On the contrary, the ηPL of 2CzPN- and DMAC-DPS-based films sharply decreases to the minimum values at quite low y, especially for the latter, indicating the dominance of relatively distance-insensitive triplet diffusion in emission quenching. It shows that ηPL variation is directly related to ϕRISC of these blue TADF emitters.

Fig. 2. Photophysical properties of white hyperfluorescence films. PL spectra of DBFDPO:40% ptBCzPO2TPTZ:y% TBRb (A), DBFDPO:40% 2CzPN:y% TBRb (B), and DBFDPO:40% DMAC-DPS:y% TBRb (C) films (vacuum evaporated, 100 nm; y = 0 to 1.0%). (D) Time-resolved emission spectra of DBFDPO:40% blue TADF emitters:y% TBRb and ptBCzPO2TPTZ:0.1% TBRb films. Blue TADF emitters are ptBCzPO2TPTZ, 2CzPN, and DMAC-DPS, respectively. (E) TADF parameter variation after TBRb doping. krS is singlet radiative rate constant. y = 0 for blue or 0.1 for white.

At extremely low TBRb concentration (y = 0.1), the yellow emission intensities are independent of blue TADF emitters (Fig. 2A to C and Fig. S2). The ~100% ηPL of the DBFDPO:40% ptBCzPO2TPTZ:0.1% TBRb film is a combined result of 100% FRET efficiency, RISC enhancement, and triplet quenching suppression. In contrast, 2CzPN- and DMAC-DPS-based films show slightly changed PL spectra with sharply decreased ηPL, reflecting exciton allocation-induced triplet nonradiation. Since triplet diffusion rather than triplet DET is predominant at y = 0.1, it suggests that triplet diffusion-induced quenching is negligible for ptBCzPO2TPTZ-based films, but significant in 2CzPN- and DMAC-DPS-based films. Furthermore, along with y increasing, yellow emissions of DMAC-DPS-based films rapidly increase, which are largely stronger than those of 2CzPN-based films, while ptBCzPO2TPTZ-based films display the weakest yellow emissions. Since FRET is a dipole–dipole resonance-based long-range energy transfer, these 3 blue TADF emitters are comparable in the FRET process. The large PL spectra difference of the dually doped films is mainly caused by intermolecular interaction-based exciton/charge exchange, namely, singlet DET. The combined analysis of ηPL and PL variations suggests that ηPL and yellow intensity are simultaneously related to the intermolecular interaction intensity in the dually doped films. Symmetrically bar-shaped DMAC-DPS has the strongest intermolecular interactions. At the same time, the highest occupied molecule orbital and lowest unoccupied molecule orbital energy levels of TBRb are respectively shallower and deeper than those of DMAC-DPS, leading to direct charge/exciton capture by TBRb (Fig. 3A). Consequently, singlet and triplet diffusions in DMAC-DPS-based films are the strongest, giving rise to the most marked yellow emissions and the lowest ηPL. On the contrary, the asymmetric and sphere-like configuration of 2CzPN somewhat reduces intermolecular interactions and limits singlet/triplet diffusions and DET, rendering the weaker yellow emissions but the higher ηPL values of 2CzPN-based films, which is similar to our previously reported all-TADF systems [29]. In comparison, ptBCzPO2TPTZ with large steric hindrance further restrains intermolecular interactions and singlet/triplet diffusion in its films. Triplet DET becomes the main channel of emission quenching, which is also limited but still appreciable at high x and y. Moreover, for ptBCzPO2TPTZ-based films, compared to DBFDPO, another conventional host, bis{2-[di(phenyl)phosphino]-phenyl}ether oxide (DPEPO), with a larger steric hindrance further decreases yellow emission intensity and mitigates ηPL reduction at high y (Fig. S3). On the contrary, without host, ηPL reduction is accelerated (Fig. S4).

Fig. 3. OLED performance. (A) Device structure and chemical structures of used materials. (B) Electroluminescence (EL) spectra at 1,000 nits. (C) Commission Internationale de l'Eclairage (CIE) coordinates and correlated color temperature (CCT) values under different doping concentrations at 1,000 nits. (D) Luminance–voltage–current density (J) relationships. (E) Efficiency–luminance correlations. CE, PE, and EQE refer to current efficiency, power efficiency, and external quantum efficiency, respectively. PE and EQE values at 100 and 1,000 cd m−2 were highlighted with arrows.

Time-resolved emission properties

Transient emission spectra (TRES) show that compared to blue-emitting 2CzPN and DMAC-DPS films, 0.1% TBRb induces marked delayed fluorescence (DF) quenching (Fig. 2D), and DF intensities and lifetimes are approximately independent of TBRb concentration (0.1% to 1.0%) (Fig. S5). This result suggests that DF quenching in 2CzPN and DMAC-DPS films is primarily due to triplet diffusion, which is consistent with their ηPL variations. In contrast, the smallest knrT makes the triplet state of ptBCzPO2TPTZ much less sensitive to TBRb doping; therefore, TRES contours of DBFDPO:x% ptBCzPO2TPTZ:y% TBRb films (x = 40%; y = 0 and 0.1) are nearly identical, despite the gradual DF reduction at y > 0.1 (Fig. S6). In the DPEPO matrix, the DF variation of ptBCzPO2TPTZ is similar (Fig. S7). At y = 0.1, when x ≤ 40, DF intensities and lifetimes remain stable, but prompt fluorescence (PF) intensity is directly proportional to x. It means that in this x range, triplet quenching can be effectively suppressed, while increasing x reduces ptBCzPO2TPTZ–TBRb distance, thus enhancing FRET. At an x range of 40 to 100, different to slightly changed PF, DF intensity and lifetime sharply decrease from x = 40 to 60 and further gradually decline until x = 100. Actually, at x = 100, the PF and DF properties of the binary films are independent on y. Therefore, when ptBCzPO2TPTZ becomes the majority (x > 50), DET via direct interactions between ptBCzPO2TPTZ and TBRb is still inevitable, rendering triplet quenching. Nevertheless, the critical x% reaching 40% still demonstrates that ~100% ϕRISC and the largely reduced knrT of ptBCzPO2TPTZ indeed effectively alleviate triplet quenching by TBRb, which establishes the feasibility of balancing FRET and triplet exclusion for optimal exciton allocation.

PF time decay curves of singly doped blue-emitting films are nearly identical to those of blue PF emissions from 0.1% TBRb codoped films (Fig. 2D and Fig. S8A). Time decay curves of yellow PF and DF are consistent with those of the corresponding blue PF and DF, respectively. PF lifetimes of TBRb in dually doped films are markedly larger than those of the DBFDPO:5% TBRb film (Fig. S8B). Therefore, yellow emissions of TBRb in these films are mainly ascribed to FRET from blue TADF emitters. However, 0.1% TBRb significantly shortens blue DF lifetimes of 2CzPN/DMAC-DPS-based films. In contrast, for DBFDPO:x% ptBCzPO2TPTZ:y% TBRb films, at x = 40, blue DF time decay curves overlap for y = 0 and 0.1, owing to effectively suppressed triplet diffusion (Fig. S9). Further increasing y of ptBCzPO2TPTZ-based films hardly changes blue PF lifetimes, but gradually reduces blue DF lifetimes, reflecting intensified DET-induced triplet quenching. On the other hand, at y = 0.1, increasing x from 10 to 40 induces elongated blue PF lifetimes of ptBCzPO2TPTZ, but does not change its blue DF lifetimes (Fig. S10). In this x range, direct excitation of ptBCzPO2TPTZ’s S1 state is gradually enhanced, but triplet quenching can still be effectively controlled. However, further increasing x from 40 to 100, blue PF lifetime becomes stable, but blue DF lifetime gradually declines, reflecting marked triplet quenching. At x = 100, increasing y from 0.1 to 1.0 can still shorten blue DF lifetime, but the variation is much smaller than at x = 40. These results further suggest that for DBFDPO:x% ptBCzPO2TPTZ:y% TBRb films, x = 40 is the critical point of balancing FRET and triplet quenching, and only a considerable x or y increase leads to serious triplet quenching, indicating the predominance of triplet DET in the quenching process.

Compared to the DBFDPO:40% ptBCzPO2TPTZ film, the ϕRISC of the DBFDPO:40% ptBCzPO2TPTZ:0.1% TBRb film is further improved to 99%, and in particular, its knrT largely decreases by one order of magnitude to 3.0 × 102 s−1 (Fig. 2E and Table S1). Simultaneously, its singlet radiative rate constant (krS) significantly increases by one-third to 1.6 × 107 s−1. Consequently, 0.1% TBRb facilitates RISC and singlet radiation and mitigates triplet quenching in ptBCzPO2TPTZ-based films. In this case, triplet excitons can be completely converted to radiative singlets by ptBCzPO2TPTZ and then respectively allocated to itself and TBRb, resulting in 100% triplet harvesting. In contrast, 0.1% TBRb induces sharp ϕRISC decreases of 8% and 12% for 2CzPN- and DMAC-DPS-based films, respectively. Furthermore, the knrT of the 2CzPN-based film is doubled, but its krS is halved, in addition to 13% increased knrS. Although the knrT and knrS of the DMAC-DPS-based film are reduced by 0.1% TBRb doping, its krS decreased by 70-fold to 1.0 × 105 s−1, which was 1/10 of its knrS. Therefore, TBRb doping hinders RISC and aggravates quenching effects in 2CzPN- and DMAC-DPS-based films. These results show that in these dually doped films, compared to ~70% ϕRISC of 2CzPN and DMAC-DPS, ptBCzPO2TPTZ with ~100% ϕRISC and small knrT establishes a basis for TBRb to largely enhance radiative transitions and reduce nonradiation. Notably, the TADF parameters of the ptBCzPO2TPTZ:0.1% TBRb film are comparable to those of the DBFDPO:40% 2CzPN:0.1% TBRb film. It suggests that despite effectively limited triplet diffusion, triplet DET to FL dopants can still cause serious quenching of TADF sensitizers.

EL performance

White OLEDs were fabricated by vacuum evaporation, following the simple 3-layer architecture (Fig. 3A). Single emissive layers DBFDPO:x% blue TADF emitter:y% TBRb were employed, in addition to using N,N′-dicarbazolyl-3,5-benzene and 2,4,6-Tris(4-(diphenylphosphoryl)phenyl)-1,3,5-triazine as hole and electron transporting layers, respectively. Concentrations of x and y were carefully explored to determine the optimal parameters (Figs. S11 to S16 and Table S2). Simultaneously, in accord with PL spectra, increasing x and y enhanced yellow emissions in EL spectra, as a result of FRET improvement. Compared to 2CzPN and DMAC-DPS with stronger intermolecular interactions, ptBCzPO2TPTZ-based devices displayed dual-peak complementary white emissions with largely higher color purities (Fig. 3B). It is noted that at y = 0.1, when x ≥ 60, the EL spectra of ptBCzPO2TPTZ-based devices almost overlapped. It indicates that at x ≥ 60, the FRET process tended to be stable, and triplet DET became dominant, leading to unchanged exciton allocation ratios for blue and yellow emissions. In contrast, at x = 40, along with y increasing from 0.1 to 1.0, EL correlated color temperature changed from 6,225 K (y = 0.1), 4,914 K (y = 0.5), to 3,921 K (y = 1.0) (Fig. 3C and Fig. S17), which are respectively close to correlated color temperature values of standard illuminants D65 (6,500 K, artificial daylight), D50 (5,000 K, simulated sunlight), and A (2,856 K, incandescent light).

It is shown that at x = 40, ptBCzPO2TPTZ-based devices realized the best performances (Figs. S11 and S14). In the y range of 0.1 to 1.0, besides luminescence beyond 30,000 cd m−2, the driving voltages at 1, 100, and 1,000 cd m−2 were as low as ~2.5, ~3.3, and ~4.1 V, respectively (Fig. 3D and Fig. S14), which are the lowest values reported so far among white hyperfluorescence OLEDs [37,43] and also comparable to the best results from all kinds of white EL devices [44] (Table S3). Moreover, the driving voltages were roughly inversely proportional to x. Obviously, at high doping concentrations, ptBCzPO2TPTZ with ambipolar feature made significant contributions to carrier injection and transportation. 2CzPN- and DMAC-DPS-based devices revealed similar situations (Figs. SS12 and S13). Nevertheless, ptBCzPO2TPTZ-based host-free devices (x = 100) displayed slightly larger driving voltages than those of DBFDPO hosted analogs. Moreover, the incorporation of DPEPO with worse electrical properties increased driving voltages. Therefore, the electrical performance of the host matrix is also considerable for balancing carrier flux, while quenching suppression by the host matrix improves luminance at specific current density (J), thereby, in turn, reducing driving voltages. Furthermore, the incorporation of TBRb largely increased the maximum luminance, since FRET to TBRb facilitates RISC and restrains triplet concentration quenching.

For DBFDPO:x% ptBCzPO2TPTZ:y% TBRb-based devices, at y = 0.1, the maximum ηEQE markedly increased from 9.6% at x = 10 to the highest value of 30.7% at x = 40 and then decreased to 11.8% at x = 100 (Figs. S11 and S15). Using the DPEPO matrix can still achieve the maximum ηEQE at x = 40 but reduces ηEQE decrease at x = 60 (Fig. S16). The optimal x for 2CzPN-based devices was also 40 (Fig. S12), different to that (10) for DMAC-DPS-based analogs (Fig. S13). At x = 40, the maximum ηEQE of ptBCzPO2TPTZ-based devices was inversely proportional to y, decreasing to 21.6% at y = 1.0. Thus, the key factor inducing ηEQE reduction is triplet quenching, including blue TADF emitter-based triplet–triplet and singlet–triplet annihilations and TBRb-induced triplet quenching. Nevertheless, ηEQE was more sensitive to x, reflecting the predominance of ptBCzPO2TPTZ in exciton utilization (Fig. S18). At x =40 and y = 0.1, the maximum ηEQE of ptBCzPO2TPTZ-based white OLEDs (30.7%) was 2.5-fold of that of 2CzPN-based devices (12.4%) and 6.3-fold of that of DMAC-DPS-based analogs (4.9%) (Fig. 3E). ηEQE values of ptBCzPO2TPTZ:y% TBRb-based devices were nearly equal to those of 2CzPN (x = 40)-based ternary devices. These results suggest that triplet DET and diffusion to TBRb are the main channels of triplet quenching. Owing to low driving voltages, ptBCzPO2TPTZ endowed its white hyperfluorescence OLEDs with a new record ηPE of 120.2 lm W−1 for the maximum. At 100 and 1,000 nits, ηPE remained 91.1 and 65.3 lm W−1, corresponding to ηEQE values of 30.0% and 26.9%, respectively. These efficiency values are far beyond previous reports of white hyperfluorescence OLEDs [37,43] and also superior to all kinds of white EL devices [44] (Table S3).

Considering a singlet ratio of 25% and an outcoupling ratio of 20% to 30% for indium tin oxide (ITO) glass, an ηEQE of ~5% corresponds to 75% to 100% singlet utilization. Therefore, for a DMAC-DPS-based device with a maximum ηEQE of ~5%, DF was almost negligible. In contrast, ptBCzPO2TPTZ- and 2CzPN-based devices exhibited ~100% and 50% singlet and triplet utilization, respectively. Furthermore, EL TRES shows that the DF intensity of the DBFDPO:40% ptBCzPO2TPTZ:0.1% TBRb-based device was markedly higher than that of the 2CzPN-based analog (Fig. 4A). The blue DF EL lifetime of ptBCzPO2TPTZ-based devices was largely longer than that of the 2CzPN-based analog, manifesting the reduced triplet quenching on ptBCzPO2TPTZ. The situation of yellow DF EL lifetime was similar. It is noted that within 150 μs, the yellow intensity of the 2CzPN-based device was stronger than that of the ptBCzPO2TPTZ-based analog. However, after 150 μs, the yellow DF intensity of the former was, in turn, smaller than that of the latter. This variation is consistent with faster triplet DET and simultaneously stronger triplet quenching in the former.

Fig. 4. EL kinetics of white hyperfluorescence OLEDs. (A) Time-resolved EL emission spectra (top) and EL time decays at blue and yellow peak wavelengths (bottom) of devices with emissive layers of DBFDPO:40% blue TADF emitter:0.1% TBRb at 1,000 nits. (B) Relationship between singlet (NS, round symbols) and triplet (NT, triangle symbols) exciton densities and J in white OLEDs and blue diodes of DBFDPO:40% ptBCzPO2TPTZ. (C) Comparison on rate coefficients of singlet–triplet (STA, kSTA) and triplet–triplet annihilations (TTA, kTTA) for the devices. a.u., arbitrary units.

Singlet (NS) and triplet (NT) exciton densities of DBFDPO:40% Blue TADF emitters:0.1% TBRb devices can be calculated as a function of current density (J)(18) (Fig. 4B):dNSdt=−(krS+kISC)NS+kRISCNT−kSTANSNT+αkTTANT2+J4de(1)

dNTdt=−(knrT+kRISC)NT+kISCNS−(1+α)kTTANT2+3J4de(2)

in which kISC and kRISC are rate constants of ISC and RISC, and d and e are recombination zone thickness and electron charge, respectively. kSTA and kTTA are rate constants of singlet–triplet and triplet–triplet annihilations, respectively. The NS and NT values were in the ranges of 1011 to 1017 and 1014 to 1019, respectively. It is shown that the NS and NT of DBFDPO:40% ptBCzPO2TPTZ:0.1% TBRb are almost the same as those of DBFDPO:40% ptBCzPO2TPTZ, manifesting the predominant roles of ptBCzPO2TPTZ in exciton conversion and utilization. Notably, kSTA and kTTA values of ptBCzPO2TPTZ-based devices reaching the levels of 10−10 and 10−15 s−1 are comparable to those of FL OLEDs [45] but markedly higher than those of common TADF diodes [18] (Fig. 4C). kSTA and kTTA are directly proportional to the oscillator strength of an exciton utilizer; therefore, it demonstrates that despite its charge transfer featured excited state, ptBCzPO2TPTZ (ηPL = 94%) is equal to locally excited TBRb (ηPL = 84%) in terms of exciton radiation, resulting in the balance of blue and yellow efficiencies for the state-of-the-art white OLEDs. In contrast to 2CzPN- and DMAC-DPS-based analogs, ptBCzPO2TPTZ significantly reduces NS by about 30- and 400-fold, and NT by 4- and 5-fold in its white OLEDs, respectively, and increases kSTA and kTTA by more than 10- and 90-fold, respectively. Thus, blue TADF emitters are dominant in exciton kinetic processes of these hyperfluorescence white diodes. Undoubtedly, compared to 2CzPN and DMAC-DPS, the superiority of ptBCzPO2TPTZ in radiation facilitation and nonradiation suppression is the primary reason for the state-of-the-art performance of its white FL OLED.

Discussion

We have demonstrated new record EL efficiencies of 30.7% for ηEQE and 120.2 lm W−1 for ηPE with a single-emissive layer white FL OLED, indicating the potential of pure organic OLEDs for large-scale daily lighting application. It is shown that 100% exciton utilization can be readily realized by hyperfluorescence white-emitting systems, when triplet DET and diffusion-induced quenching can be effectively suppressed. The fundamental solution is to reduce triplet concentration in the device, which requires TADF sensitizers to have ~100% ϕRISC and low enough knrT. Simultaneously, TADF sensitizers with a large steric hindrance are advantageous in reducing triplet diffusion to FL emitters. It is noted that FRET from eligible TADF materials, like ptBCzPO2TPTZ herein, to FL emitter can further improve RISC and singlet radiation and alleviate triplet nonradiation. These results provide new insights into optimizing exciton allocation and utilization in TADF-based systems. We obtained a power efficiency of 65.3 lm W−1 at 1,000 cd m−2, which can be doubled to over 130 lm W−1 by additional outcoupling enhancement, thereby surpassing the current benchmarks of daily lightings (100 lm W−1 for white LED and 70 lm W−1 for FL tube). For large-scale applications, the lifetime issue of blue TADF emitters should be solved in the future. Nevertheless, our results suggest that pure organic white OLEDs hold the promise of satisfying all demands for ideal lighting sources, including high efficiency, low cost, ecofriendliness, and so on.

Materials and Methods

Transient emission measurement

The films for measurement were prepared by vacuum evaporation. PL TRES were measured by an Edinburgh FLS 1000 fluorescence spectrophotometer using a time-correlated single photon counting method with a nanosecond and a microsecond pulsed light source for 100 ps to 10 s lifetime measurement, a synchronization photomultiplier for signal collection, and a multichannel scaling mode of the PCS900 fast counter PC plug-in card for data processing.

Fabrication and characterization of OLEDs

The ITO substrate was cleaned with detergent and deionized water, dried in an oven at 120 °C for 4 h, treated with oxygen plasma for 3 min, and then transferred to a deposition chamber. Devices were fabricated by evaporating each layer onto the ITO substrate sequentially at a pressure below 1 × 10−6 Torr. A MoO3 layer was first deposited on the ITO substrate at 0.1 nm s−1. The deposition rates of organic layers were 0.1 to 0.3 nm s−1. Then, a 1-nm layer of LiF was deposited at 0.1 nm s−1. Finally, a 100-nm-thick layer of Al was deposited at 0.6 nm s−1 as the cathode. The devices were then transferred to a glovebox and encapsulated with hot melt glue. The emission areas of the devices were 0.09 cm2. EL spectra were measured with a PR655 spectra colorimeter. Current–density–voltage and brightness–voltage curves of the devices were measured using a Keithley 4200 source meter and a calibrated silicon photodiode. All the measurements were carried out in atmosphere.

Acknowledgments

H.X. thanks the great assistance of G. Xie (Wuhan University) in electronics analysis. Funding: This study was supported by the National Natural Science Foundation of China (92061205, 62175060, 51873056, 61905070, and 22005088), the Young Innovative Team Supporting Projects of Heilongjiang Province, the National Science Fund for Excellent Young Scholars of Heilongjiang Province (YQ2020B006), the Postdoctoral Science Foundation of Heilongjiang Province (LBH-Q2116), and the Fund for Distinguished Young Scholars of Heilongjiang University (JCL202001). Author contributions: H.X. conceived the idea and directed the project. D.D. and C.D. fabricated and tested OLEDs. Y.W. performed theoretical calculations. Z.W. synthesized ptBCzPO2TPTZ. H.X., D.D., Y.W., C.H., J.Z., and S.C. analyzed data. H.X., D.D., and Y.W. wrote the manuscript. Competing interests: The authors declare that they have no competing interests.

Data Availability

All other data are available from the authors upon reasonable request.

Supplementary Materials

Supplementary Materials Experimental Details

Fig. S1. Photoluminescence (PL) spectra and photoluminescence quantum yields (PLQY, ηPL) of DBFDPO:x% ptBCzPO2TPTZ:0.1% TBRb films at different x.

Fig. S2. Comparison on PL spectra of blue TADF emitter-based films before and after doping 0.1% TBRb. Bis{2-[di(phenyl)phosphino]-phenyl}ether oxide (DPEPO) and DBFDPO are used as host matrixes.

Fig. S3. PL spectra and ηPL of DBFDPO:40% ptBCzPO2TPTZ:y% TBRb films at different y.

Fig. S4. PL spectra and ηPL of ptBCzPO2TPTZ:y% TBRb films at different y.

Fig. S5. TRES spectra of DBFDPO:40% 2CzPN:y% TBRb (above) and DBFDPO:40% DMAC-DPS:y% TBRb (below) films at different y.

Fig. S6. TRES spectra of DBFDPO:x% ptBCzPO2TPTZ:y% TBRb films at different x and y.

Fig. S7. TRES spectra of DPEPO:40% ptBCzPO2TPTZ:y% TBRb films at different y.

Fig. S8. Prompt fluorescence (PF, insets) and delayed fluorescence (DF) time decay curves of (A) blue and (B) yellow components for blue TADF emitter-based films and the DBFDPO:5% TBRb film for comparison.

Fig. S9. Prompt fluorescence (PF, insets) and delayed fluorescence (DF) time decay curves of (A) blue and (B) yellow components for DBFDPO:40% ptBCzPO2TPTZ:y% TBRb films at different y. Yellow DF lifetime variation is consistent with blue DF, but yellow PF lifetimes are slightly increased, due to the direct excitation and radiation of TBRb.

Fig. S10. Prompt fluorescence (PF, insets) and delayed fluorescence (DF) time decay curves of (A) blue and (B) yellow components for DBFDPO:x% ptBCzPO2TPTZ:0.1% TBRb films at different x.

Fig. S11. EL performance of DBFDPO:x% ptBCzPO2TPTZ:0.1% TBRb-based devices. (A) Current density (J)–voltage–luminance relationship and EL spectra at 1,000 nits (inset). (B) Efficiencies vs. luminance correlations.

Fig. S12. EL performance of DBFDPO:x% 2CzPN:0.1% TBRb-based devices. (A) Current density (J)–voltage–luminance relationship and EL spectra at 1,000 nits (inset). (B) Efficiencies vs. luminance correlations.

Fig. S13. EL performance of DBFDPO:x% DMAC-DPS:0.1% TBRb-based devices. (A) Current density (J)–voltage–luminance relationship and EL spectra at 1,000 nits (inset). (B) Efficiencies vs. luminance correlations.

Fig. S14. EL performance of DBFDPO:40% ptBCzPO2TPTZ:y% TBRb-based devices. (A) Current density (J)–voltage–luminance relationship and EL spectra at 1,000 nits (inset). (B) Efficiencies vs. luminance correlations.

Fig. S15. EL performance of ptBCzPO2TPTZ:y% TBRb-based devices. (A) Current density (J)–voltage–luminance relationship and EL spectra at 1,000 nits (inset). (B) Efficiencies vs. luminance correlations.

Fig. S16. EL performance of DPEPO:x% ptBCzPO2TPTZ:0.1% TBRb-based devices. (A) Current density (J)–voltage–luminance relationship and EL spectra at 1,000 nits (inset). (B) Efficiencies vs. luminance correlations.

Fig. S17. Chromatic properties of DBFDPO:40% ptBCzPO2TPTZ:y% TBRb-based white OLEDs.

Fig. S18. Dependence of the maximum EQE values on doping concentrations of DBFDPO:x% ptBCzPO2TPTZ:y% TBRb-based devices.

Table S1. Photophysical properties of ptBCzPO2TPTZ-, 2CzPN-, and DMAC-DPS-based films.

Table S2. Device performance of DBFDPO:x% blue TADF emitter:y% TBRb ptBCzPO2TPTZ- and 2CzPN-based white OLEDs.

Table S3. EL performance of representative TADF-involved white OLEDs.
==== Refs
References

1. Kido J, Kimura M, Nagai K. Multilayer white light-emitting organic electroluminescent device. Science. 1995;267 (5202 ):1332–1334.17812607
2. Nishide J-i, Nakanotani H, Hiraga Y, Adachi C. High-efficiency white organic light-emitting diodes using thermally activated delayed fluorescence. Appl Phys Lett. 2014;104 (23 ):233304.
3. Byeon SY, Lee DR, Yook KS, Lee JY. Recent progress of singlet-exciton-harvesting fluorescent organic light-emitting diodes by energy transfer processes. Adv Mater. 2019;31 (34 ):1803714.
4. Baldo MA, O’Brien DF, You Y, Shoustikov A, Sibley S, Thompson ME, Forrest SR. Highly efficient phosphorescent emission from organic electroluminescent device. Nature. 1998;395 (6698 ):151–154.
5. Uoyama H, Goushi K, Shizu K, Nomura H, Adachi C. Highly efficient organic light-emitting diodes from delayed fluorescence. Nature. 2012;492 (7428 ):234–238.23235877
6. Liu Y, Li C, Ren Z, Yan S, Bryce MR. All-organic thermally activated delayed fluorescence materials for organic light-emitting diodes. Nat Rev Mater. 2018;3 (4 ):18020.
7. Yang Z, Mao Z, Xie Z, Zhang Y, Liu S, Zhao J, Xu J, Chi Z, Aldred MP. Recent advances in organic thermally activated delayed fluorescence materials. Chem Soc Rev. 2017;46 (3 ):915–1016.28117864
8. Wong MY, Zysman-Colman E. Purely organic thermally activated delayed fluorescence materials for organic light-emitting diodes. Adv Mater. 2017;29 (22 ):160544.
9. Tao Y, Yuan K, Chen T, Xu P, Li H, Chen R, Zheng C, Zhang L, Huang W. Thermally activated delayed fluorescence materials towards the breakthrough of organoelectronics. Adv Mater. 2014;26 (47 ):7931–7958.25230116
10. Samanta PK, Kim D, Coropceanu V, Brédas JL. Up-conversion intersystem crossing rates in organic emitters for thermally activated delayed fluorescence: Impact of the nature of singlet vs triplet excited states. J Am Chem Soc. 2017;139 (11 ):4042–4051.28244314
11. Penfold TJ , Dias FB , Monkman AP. The theory of thermally activated delayed fluorescence for organic light emitting diodes. Chem Commun. 2018;54 (32 ):3926–3935.
12. Tang X, Cui L-S, Li H-C, Gillett AJ, Auras F, Qu Y-K, Zhong C, Jones STE, Jiang Z-Q, Friend RH, et al. Highly efficient luminescence from space-confined charge-transfer emitters. Nat Mater. 2020;19 (12 ):1332–1338.32541938
13. Kondo Y, Yoshiura K, Kitera S, Nishi H, Oda S, Gotoh H, Sasada Y, Yanai M, Hatakeyama T. Narrowband deep-blue organic light-emitting diode featuring an organoboron-based emitter. Nat Photon. 2019;13 (10 ):540–546.
14. Jeon SO, Lee KH, Kim JS, Ihn S-G, Chung YS, Kim JW, Lee H, Kim S, Choi H, Lee JY. High-efficiency, long-lifetime deep-blue organic light-emitting diodes. Nat Photonics. 2021;15 (3 ):208–215.
15. Xiang Y, Li P, Gong S, Huang Y-H, Wang C-Y, Zhong C, Zeng W, Chen Z, Lee W-K, Yin X, et al. Acceptor plane expansion enhances horizontal orientation of thermally activated delayed fluorescence emitters. Sci Adv. 2020;6 (41 ):eaba7855.33036963
16. Wu T-L, Huang M-J, Lin C-C, Huang P-Y, Chou T-Y, Chen-Cheng R-W, Lin H-W, Liu R-S, Cheng C-H. Diboron compound-based organic light-emitting diodes with high efficiency and reduced efficiency roll-off. Nat Photon. 2018;12 (4 ):235–240.
17. Wolf MW, Legg KD, Brown RE, Singer LA, Parks JH. Photophysical studies on the benzophenones. Prompt and delayed fluorescences and self-quenching. J Am Chem Soc. 1975;97 (16 ):4490–4497.
18. Masui K, Nakanotani H, Adachi C. Analysis of exciton annihilation in high-efficiency sky-blue organic light-emitting diodes with thermally activated delayed fluorescence. Org Electron. 2013;14 (11 ):2721–2726.
19. Zhang Q, Li B, Huang S, Nomura H, Tanaka H, Adachi C. Efficient blue organic light-emitting diodes employing thermally activated delayed fluorescence. Nat Photon. 2014;8 (4 ):326–332.
20. Yao L, Yang B, Ma Y. Progress in next-generation organic electroluminescent materials: Material design beyond exciton statistics. Sci China Chem. 2014;57 (3 ):335–345.
21. Tao Y, Xu L, Zhang Z, Chen R, Li H, Xu H, Zheng C, Huang W. Achieving optimal self-adaptivity for dynamic tuning of organic semiconductors through resonance engineering. J Am Chem Soc. 2016;138 (30 ):9655–9662.27403886
22. Ahn DH, Kim SW, Lee H, Ko IJ, Karthik D, Lee JY, Kwon JH. Highly efficient blue thermally activated delayed fluorescence emitters based on symmetrical and rigid oxygen-bridged boron acceptors. Nat Photon. 2019;13 (8 ):540–546.
23. Yang H, Liang Q, Han C, Zhang J, Xu H. A phosphanthrene oxide host with close sphere packing for ultralow-voltage-driven efficient blue thermally activated delayed fluorescence diodes. Adv Mater. 2017;29 (38 ):1700553.
24. Han C, Zhang Z, Ding D, Xu H. Dipole-dipole interaction management for efficient blue thermally activated delayed fluorescence diodes. Chem. 2018;4 (9 ):2154–2167.
25. Han C, Duan C, Yang W, Xie M, Xu H. Allochroic thermally activated delayed fluorescence diodes through field-induced solvatochromic effect. Sci Adv. 2017;3 (9 ):e1700904.28929136
26. Liang J, Li C, Zhuang X, Ye K, Liu Y, Wang Y. Novel blue bipolar thermally activated delayed fluorescence material as host emitter for high-efficiency hybrid warm-white OLEDs with stable high color-rendering index. Adv Funct Mater. 2018;28 (17 ):1707002.
27. Zeng J, Guo J, Liu H, Zhao Z, Tang BZ. A multifunctional bipolar luminogen with delayed fluorescence for high-performance monochromatic and color-stable warm-white OLEDs. Adv Funct Mater . 2020;30 (17 ):2000019.
28. Chunbo D, Chunmiao H, Ruiming D, Ying W, Hui X. High-efficiency blue dual-emissive exciplex boosts full-radiative white electroluminescence. Adv Opt Mater. 2018;6 (19 ):1800437.
29. Han C, Du R, Xu H, Han S, Ma P, Bian J, Duan C, Wei Y, Sun M, Liu X, et al. Ladder-like energy-relaying exciplex enables 100% internal quantum efficiency of white TADF-based diodes in a single emissive layer. Nat Commun. 2021;12 (1 ):3640.34131127
30. Wu Z, Liu Y, Yu L, Zhao C, Yang D, Qiao X, Chen J, Yang C, Kleemann H, Leo K, et al. Strategic-tuning of radiative excitons for efficient and stable fluorescent white organic light-emitting diodes. Nat Commun. 2019;10 (1 ):2380.31147542
31. Zhang C, Zhang D, Bin Z, Liu Z, Zhang Y, Lee H, Kwon JH, Duan L. Color-tunable all-fluorescent white organic light-emitting diodes with a high external quantum efficiency over 30% and extended device lifetime. Adv Mater. 2022;34 (22 ):2103102.
32. Sun J, Zhang J, Liang Q, Wei Y, Duan C, Han C, Xu H. Charge-transfer exciton manipulation based on hydrogen bond for efficient white thermally activated delayed fluorescence. Adv Funct Mater. 2020;30 (9 ):1908568.
33. Li J, Ding D, Tao Y, Wei Y, Chen R, Xie L, Huang W, Xu H. A significantly twisted spirocyclic phosphine oxide as a universal host for high-efficiency full-color thermally activated delayed fluorescence diodes. Adv Mater. 2016;28 (16 ):3122–3130.26923460
34. Li W, Li B, Cai X, Gan L, Xu Z, Li W, Liu K, Chen D, Su S-J. Tri-spiral donor for high efficiency and versatile blue thermally activated delayed fluorescence materials. Angew Chem Int Ed. 2019;58 (33 ):11301–11305.
35. Li C, Duan C, Han C, Xu H. Secondary acceptor optimization for full-exciton radiation: Toward sky-blue thermally activated delayed fluorescence diodes with external quantum efficiency of ≈30%. Adv Mater. 2018;30 (50 ):1804228.
36. Ding D, Wang Z, Li C, Zhang J, Duan C, Wei Y, Xu H. Highly efficient and color-stable thermally activated delayed fluorescence white light-emitting diodes featured with single-doped single emissive layers. Adv Mater. 2020;32 (10 ):e1906950.31990429
37. Cui L-S, Gillett AJ, Zhang S-F, Ye H, Liu Y, Chen X-K, Lin Z-S, Evans EW, Myers WK, Ronson TK, et al. Fast spin-flip enables efficient and stable organic electroluminescence from charge-transfer states. Nat Photon. 2020;14 (10 ):636–642.
38. Nakanotani H, Higuchi T, Furukawa T, Masui K, Morimoto K, Numata M, Tanaka H, Sagara Y, Yasuda T, Adachi C. High-efficiency organic light-emitting diodes with fluorescent emitters. Nat Commun. 2014;5 (1 ):4016.24874292
39. Zhang D, Duan L, Li C, Li Y, Li H, Zhang D, Qiu Y. High-efficiency fluorescent organic light-emitting devices using sensitizing hosts with a small singlet–triplet exchange energy. Adv Mater. 2014;26 (29 ):5050–5055.24944186
40. Gottardi S, Barbry M, Coehoorn R, van Eersel H. Efficiency loss processes in hyperfluorescent OLEDs: A kinetic Monte Carlo study. Appl Phys Lett. 2019;114 (7 ):073301.
41. Duan C, Xin Y, Wang Z, Zhang J, Han C, Xu H. High-efficiency hyperfluorescent white light-emitting diodes based on high-concentration-doped TADF sensitizer matrices via spatial and energy gap effects. Chem Sci. 2022;13 (1 ):159–169.
42. Wu C, Djurovich PI, Thompson ME. Study of energy transfer and triplet exciton diffusion in hole-transporting host materials. Adv Funct Mater. 2009;19 (19 ):3157–3164.
43. Zhang C, Lu Y, Liu Z, Zhang Y, Wang X, Zhang D, Duan L. A π–D and π–A exciplex-forming host for high-efficiency and long-lifetime single-emissive-layer fluorescent white organic light-emitting diodes. Adv Mater. 2020;32 (42 ):2004040.
44. Xiang H, Wang R, Chen J, Li F, Zeng H. Research progress of full electroluminescent white light-emitting diodes based on a single emissive layer. Light Sci Appl. 2021;10 (1 ):206.34611124
45. Kasemann D, Brückner R, Fröb H, Leo K. Organic light-emitting diodes under high currents explored by transient electroluminescence on the nanosecond scale. Phys Rev B. 2011;84 (11 ):115208.
