English

• Thanks to the distinct features of self-luminescence, flexibility, and high contrast, organic light-emitting diodes (OLEDs) have been widely applied in the fields of consumer electronic displays [1,2], augmented/virtual reality [3,4], biomedicine [5,6], and so on. As the third-generation emissive materials, thermally activated delayed fluorescence (TADF) emitters are at the forefront of OLED research because they are free of noble metals and can harvest 100% exciton for emission due to up-conversion from the lowest triplet (T₁) to singlet (S₁) excited states via reverse intersystem crossing (RISC) mechanism [7,8]. The key to this process is a small energy gap between the S₁ and T₁ states (∆E_ST). Achieving a small ∆E_ST without sacrificing photoluminescence (PL) efficiency is a challenge. The classical approach to this challenge involves constructing donor-acceptor (D-A) type molecules [9,10]. Since Adachi's pioneering work [11], there has been an explosion of innovation, leading to the development of numerous highly efficient TADF emitters and significant advancements in the field [12-17].

  Carbazole-benzonitrile derivatives (CzBNs) stand out as a class of exemplary TADF emitters. Some of the CzBNs emitters feature a single type of donor, termed as homo-donor structures [18-20], and some are hetero-donor structures, which incorporate two or more distinct donor groups [21-25]. The latter has garnered increasing interest due to their potential to reduce the delayed lifetime and improve device stability. Despite the impressive efficiency and stability of CzBNs, TADF OLEDs based on these materials often encounter a significant efficiency roll-off, which is primarily attributed to exciton quenching [26-29]. To overcome this challenge and unlock the full potential of CzBNs in high-performance OLEDs, a rational molecular design is essentially required.

  Up until now, the majority of CzBN-based TADF molecules have featured symmetric rigid structures [18-20], while asymmetric rigid structures have been less explored [22]. In this contribution, we introduce a novel approach by designing a pair of hetero-donor-based asymmetric rigid CzBN molecules. These molecules are characterized by alternating donor and acceptor moieties at their periphery. Specifically, carbazole (Cz) and modified-Cz moieties groups serve as donors, while cyano (CN) and fluorine (F) units act as acceptors. This configuration offers several advantages: (1) The inclusion of a second or third donor enhances the interplay between charge transfer (CT) and localized excited (LE) states, thereby promoting RISC [21,24,30]. (2) The linear arrangement of the D-A-D system facilitates the formation of delocalized excited states and alleviates structural relaxation, and further boosts RISC and luminous efficiency [30]. (3) The asymmetric rigid structures deliver significant steric hindrance, which minimizes molecular vibration and suppresses non-radiative deactivation of excitons. Based on such a motif, (2S,3R,4R,6R)–4-(9H-carbazol-9-yl)–2,3,6-tri(3,6-diphenyl-9H-carbazol-9-yl)–5-fluoro-bezonitrile (referred as 3DPh) and (2R,4R,5R,6S)–4-(9H-carbazol-9-yl)–2,6-bis(3,6-diphenyl-9H-carbazol-9-yl)–3-fluoro-5-(3-phenyl-9H-carbazol-9-yl) benzonitrile (referred as 2DPh) were synthesized here. Detailed synthesis procedures and nuclear magnetic resonance (NMR) characterizations are presented in Scheme S1 and Figs. S1-S4 (Supporting information).

  To get insight into molecular structural properties, the interior structures of the compounds were visualized by using X-ray single crystal diffraction (SC-XRD). The SC-XRD results reveal that the new compounds consist of a central benzene ring, with Cz and phenyl-modified Cz groups positioned at the periphery of this central benzene ring (Fig. 1). The three-dimensional spatial conformation and primary crystallographic data of these compounds are further detailed in Fig. S5 and Table S1 (Supporting information). The observations indicate that the two individual crystals are part of asymmetric polyaromatic ring structures, providing valuable insight into the molecular architecture and potential implications for their physical and chemical properties.

  Figure 1

  

  Figure 1.  X-ray single crystal diffraction structures of the two compounds: (left) 3DPh and (right) 2DPh.

  DownLoad: Full-Size Img PowerPoint

  Both the compounds show similar absorption spectra, as shown in Figs. 2a and b. The absorption peaked at around 290 nm is attributed to the π−π* transition of Cz or modified-Cz donors. Additionally, the relatively weak and broad absorption bands in the range of 330–450 nm are indicative of intramolecular charge transfer (ICT) transitions from DPhCz/PhCz donors to BN acceptors. To further elucidate the origins of these broad absorption bands, we employed Gaussian fitting to dissect the spectra between 280 nm and 500 nm into three distinct components. The decomposed curves peak approximately at 290, 350 and 380 nm, respectively, corresponding to the π−π* band of the Cz groups, the normal, and novel CT bands. Similar to the cases of 4CzBN and 5CzBN [30], the normal and novel CT bands are believed to arise from the formation of localized CT (^loCT) and delocalized CT (^deCT) states. This is because the two compounds are capable of forming a linear D-A-D structure [30], specifically with the BN unit and DPhCz groups at the 3- and 6-positions for 3DPh, BN and 2-DPhCz and 5-PhCz groups for 2DPh. The HOMO energy levels of the two compounds are determined from measuring their pristine films using ultraviolet photoelectron spectroscopy (UPS) [31,32]. As shown in Fig. 2c, the HOMO levels are found to be −5.92 eV for 3DPh and −5.87 eV for 2DPh. By combining these HOMO levels with the optical bandgaps (E_g) determined from the absorption spectra, the corresponding LUMO levels are estimated to be −2.28 eV for 3DPh and −2.18 eV for 2DPh.

  Figure 2

  

  Figure 2.  Absorption spectra with Gaussian fitting curves of (a) 3DPh and (b) 2DPh. (c) UPS spectra and (d) transient PL decays of 3DPh and 2DPh. PL spectra at 300 K, 80 K, and 80 K with 2 ms delay of (e) 3DPh and (f) 2DPh.

  DownLoad: Full-Size Img PowerPoint

  The transient PL decay profiles for the emission of both compounds in 2,2-dimethyltetrahydrofuran (2Me-THF) were monitored at 300 K. As presented in Fig. 2d, both compounds exhibit bi-exponential decay, characterized by a nano-second-scale prompt component (~10 ns) and a notably short sub-microsecond-scale delayed component (~100 ns) by fitting with a biexponential decay function, indicating highly efficient T₁-to-S₁ spin-flip transition. The steady state PL spectra of each compound were also measured in 2Me-THF solution (Figs. 2e and f). Both compounds exhibit almost identical PL spectra, with broad and structureless emissions peaking at 470 nm at 300 K, which declares CT feature of S₁ states (¹CT₁). In contrast, the PL spectra at 80 K reveal two distinct peaks at 448 and 475 nm, delivering a well-resolved vibronic structure. These two peaks are likely attributed to different excited triplet states. The peak at 448 nm might stem from ³LE state of modified-Cz moieties [21], while the peak at 475 nm should originate from the CT state. The phosphorescence emission is confirmed by the virtually identical prompt and delayed PL spectra at 80 K, manifesting that the T₁ state is a mixed state of LE and CT characters, as expected in hetero-donor systems [21,24,30]. According to El-Sayed rule [33,34], spin-orbital coupling (SOC) between S₁ and T₁ is permitted as their distinct excited state features and the change in spatial orbital angular momentum [35], which accelerates delayed fluorescence and bring about a remarkably short delayed lifetime. From the spectra, the energies of S₁ and T₁ states for the two emitters were calculated to be 2.95 and 2.77 eV, respectively, yielding a small ∆E_ST of 0.18 eV.

  Theoretical computations based on density function theory (DFT) were carried out to examine the geometrical and electronic characteristics of the two compounds. Fig. 3 shows the optimized ground state (S₀) structures using Gaussian software at the B3LYP/6–31G(d) level [36]. In both emitters, the dihedral angles between the Cz units and the central phenyl ring are notably large (>50^°), consistent with the observations from the crystal structures. Additionally, Fig. 3 also illustrates the frontier molecular orbital (FMO) distributions of their S₀ states of the compounds. The computed lowest unoccupied molecular orbitals (LUMOs) of both compounds are principally located on the BN unit, while the highest occupied molecular orbitals (HOMOs) are predominantly situated at the 2-, 3-sites or 5-, 6-sites modified-Cz moieties, which are spatially well separated from their LUMOs and weakly overlapped with LUMOs at the central benzene ring. Due to this effective spatial separation between the FMOs, the exchange integral is expected to be small, and the excitation of HOMO-to-LUMO is assumed to be an ICT excitation. It is also noteworthy that the HOMO distribution in 3DPh is more delocalized and extends considerably to the peripheral phenyl of the DPhCz group at the 3-position compared to 2DPh. This increased delocalization could imply that 3DPh may exhibit higher radiative efficiency.

  Figure 3

  

  Figure 3.  The structures of the 3DPh and 2DPh compounds (upper, the values in parentheses are crystal data), and their calculated FMO and NTO distributions via Gaussian at B3LYP/6–31G(d) level (lower).

  DownLoad: Full-Size Img PowerPoint

  Excited state (S₁ and T₁) structures and natural transition orbitals (NTOs) [37] for the two compounds were further calculated by using time-dependent DFT (TD-DFT) at a B3LYP/6–31G(d) level. Similar to S₀ state, the highest occupied NTOs of the S₁ states are distributed on 2-, 3-positions (or 5-, 6-positions) of the modified Cz units for both molecules, while the lowest unoccupied NTOs are delocalized over the BN unit (Fig. 3). Interestingly, for the T₁ states, the highest occupied NTOs are delocalized across a linearly positioned pair of modified-Cz units (i.e., 3-, 6-positions or 2-, 5-positions) along with the BN core. The S₁ and T₁ states exhibit characteristics of CT and hybrid local charge transfer (HLCT), respectively, which is in good agreement with experimental observations. A linear D-A-D structure facilitates the formation of delocalized excited state and promotes RISC [30]. The calculated energy gaps of ∆E_ST are 0.094 eV for 3DPh and 0.110 eV for 2DPh, indicating the potential for efficient TADF performance. Both emitters have similar HOMO and LUMO values, although there is a deviation between the experimental results and the computed values. A summary of the calculated and experimental parameters for the two emitters is presented in Table 1, showing a good overall consistency between the two sets of data.

  Table 1

  

  Table 1.  The experimental (Ex) and calculated (Cal) parameters of the two emitters.

  DownLoad: CSV

  The electroluminescent (EL) performance of devices incorporating 3DPh and 2DPh as the emissive materials were studied. For the host material, 1,3-bis(carbazol-9-yl)benzene (mCP) was chosen due to its high triplet energy (T₁ = 2.90 eV), which is compatible with the emitters. The molecular structures and energy level alignment of the organic materials used in this study are illustrated in Figs. 4a and b, respectively. The PL spectrum of mCP overlaps well with the absorption spectra of the two emitters (Fig. 4c), indicating efficient energy transfer from the mCP host to the emitters. Di-[4-(N,N-ditolyl-amino)-phenyl]cyclohexane (TAPC) and 1,3,5-tris(1-phenyl-1H-benzimidazol-2-yl)benzene (TPBi) were respectively selected as the hole and electron transporting layers due to their outstanding carrier mobility. The detailed device structure is ITO/MoO₃ (3 nm)/TAPC (50 nm)/mCP:emitter (x wt%, 20 nm)/TPBi (60 nm)/LiF (0.8 nm)/Al (150 nm). Here, MoO₃ and LiF serve as the hole and electron injection layers, respectively, and the doping concentration of each emitter was optimized. With a 5% increment, the doping concentration of the emitters in mCP (x%) was incrementally increased from 5% to 20%. The devices with emitters of 3DPh and 2DPh were labelled as sets A and B, respectively. The resultant device performance can be found in Figs. S6 and S7 (Supporting information), with the main parameters summarized in Tables S2 and S3 (Supporting information).

  Figure 4

  

  Figure 4.  (a) The molecular structures and (b) energy level alignment of organic semiconductors in this study. (c) PL spectrum of mCP and absorption of the 3DPh and 2DPh emitters.

  DownLoad: Full-Size Img PowerPoint

  As shown in Fig. S6, the devices based on 3DPh exhibit nearly concentration-independent current density-voltage (J–V) characteristics and maintain a consistent turn-on voltage of 4.60 ± 0.06 V (at 10 cd/m²). The luminance-voltage (L–V) profiles indicate a shift towards lower voltages with increasing doping content. The optimum 3DPh-based device with a doping content of 15% achieves a maximum luminance of 47,042 cd/m² and peak efficiencies of 37.5 cd/A, 13.0% and 21.9 lm/W. Fig. S7 illustrates that for the 2DPh-based devices, the current density increases with higher doping contents under a fixed voltage. Similarly, the L–V profiles show a reduction in driving voltage as the doping content rises, from 5.88 V for 5%-doped device to 5.13 V for the 20%-doped device at a constant luminance of 10 cd/m². The device doped with 20% 2DPh delivers the best performance, with the lowest driving voltage (5.13 V at 10 cd/m²), the maximum luminance (24,404 cd/m² at 14.7 V) and the highest efficiencies (18.0 cd/A, 9.1%, and 10.2 lm/W). One also notes that the EL spectra shown in Figs. S6d and S7d slightly redshift when the doping level rises from 5 wt% to 20 wt%, which may be attributed to the emission of aggregation states caused by stronger intermolecular interaction [38].

  Fig. 5 and Table 2 present the EL characteristics and key parameters of the devices with the optimal content of emitters, i.e., ITO/MoO₃/TAPC/mCP: 15 wt% 3DPh/TPBi/LiF/Al and ITO/MoO₃/TAPC/mCP: 20 wt% 2DPh/TPBi/LiF/Al. Because of the substituent effect of DPhCz group at the 3-position, the 3DPh-based device outperforms the 2DPh-based device in terms of current density, luminance, and efficiency (Figs. 5a and b). Both devices demonstrate a remarkable low efficiency roll-off. For the 3DPh-based device, the EQE maintains 96% and 79% of its maximum value at 1000 and 5000 cd/m², respectively (Fig. 5b). For the 2DPh-based device, the corresponding values are 87% and 65%. This suppressed roll-off is likely due to the short-delayed lifespan (100 ns), which moderates the density of T₁ states in emissive layer and alleviates the annihilation of T₁ states [8,21,39]. The reduction in donor strength of the PhCz group relative to the DPhCz gives rise to a considerable blueshift of 2DPh-based device compared to the 3DPh-based device (Fig. 5c). The 2DPh-based device exhibits a sky-blue emission with a peak wavelength of 487 nm and Commission Internationale de l´Eclairage (CIE) coordinates of (0.16, 0.31), while the 3DPh-based device has a peak emission at 506 nm with CIE coordinates of (0.23, 0.55) (Fig. 5d).

  Figure 5

  

  Figure 5.  EL characteristics of the optimized devices with the two emitters: (a) J–V–L, (b) η_CE–L–η_EQE. (c) EL spectra and (d) CIE coordinates at 1000 cd/m².

  DownLoad: Full-Size Img PowerPoint

  Table 2

  

  Table 2.  The key properties of the OLED devices with different emitters.

  DownLoad: CSV

  In conclusion, we have successfully synthesized a couple of CzBN-based emitters featuring asymmetrical rigid structures, designed with alternate donor and acceptor groups arranged in a radially D-A-D structure. These emitters are characterized by a notably short delayed lifespan (~100 ns) because of the formation of delocalized excited state induced by the linear D-A-D structure and strong SOC by distinct excited state character of S₁ and T₁, which promote T₁-to-S₁ spin-flip process. Both emitters have been utilized to fabricate efficient OLEDs, achieving an EQE of 13.0% for green emission and 9.1% for sky-blue emission. Particularly, the devices demonstrate exceptionally low efficiency roll-off, a result of the rapid T₁-to-S₁ flip process that suppresses triplet exciton annihilation. The EQE roll-offs from the maximum to 1000, 5000, and 10,000 cd/m² are correspondingly 3.8%, 21%, and 33% for the green OLED, and 13%, 35%, and 46% for the sky-blue OLED. These observations confirm the effectiveness of our molecular design strategy in enhancing the spin-flip process and improving the roll-off performance of OLEDs, showcasing the potential of these materials for high-performance display and lighting applications.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  CRediT authorship contribution statement

  Zhaoyue Lü: Writing – original draft, Supervision, Investigation, Data curation, Conceptualization. Tiantian Chai: Writing – original draft, Data curation. Yichao Jin: Writing – original draft, Data curation. Xiao Wang: Data curation. Ye Zou: Writing – review & editing, Conceptualization. Lijiang Zhang: Data curation. Jiankang Feng: Data curation. Mengtong Zhang: Data curation. Shuo Wang: Data curation. Chichong Lu: Writing – original draft, Investigation, Conceptualization. Guofan Jin: Writing – review & editing, Writing – original draft, Supervision, Investigation, Conceptualization.

  Acknowledgments

  This work was financially supported by the National Natural Science Foundation of China (Nos. T2441002 and 22175186).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2025.110817.

  
  
• 1. [1]

     Y. Lee, H. Cho, H. Yoon, et al., Adv. Mater. Technol. 8 (2023) 2201067. doi: 10.1002/admt.202201067

  2. [2]

     G. Hong, X. Gan, C. Leonhardt, et al., Adv. Mater. 33 (2021) 2005630. doi: 10.1002/adma.202005630

  3. [3]

     C.M. Kang, H. Lee, J. Inf. Disp. 23 (2022) 19–32. doi: 10.1080/15980316.2021.1917461

  4. [4]

     W.J. Joo, J. Kyoung, M. Esfandyarpour, et al., Science 370 (2020) 459–463. doi: 10.1126/science.abc8530

  5. [5]

     Y. Lee, J.W. Chung, G.H. Lee, et al., Sci. Adv. 7 (2021) eabg9180. doi: 10.1126/sciadv.abg9180

  6. [6]

     H. Chae, Y. Park, Y. Jo, et al., Adv. Mater. Interfaces 10 (2023) 2202443. doi: 10.1002/admi.202202443

  7. [7]

     X. Hu, Y. Qin, Z. Li, et al., Chin. Chem. Lett. 33 (2022) 4645–4648. doi: 10.1016/j.cclet.2021.12.047

  8. [8]

     X.C. Fan, K. Wang, Y.Z. Shi, et al., SmartMat 4 (2023) e1122. doi: 10.1002/smm2.1122

  9. [9]

     S. Hirata, Y. Sakai, K. Masui, et al., Nat. Mater. 14 (2015) 330–336. doi: 10.1038/nmat4154

  10. [10]

      Y. Mei, D. Liu, J. Li, H. Li, W. Wei, J. Mater. Chem. C 9 (2021) 5885–5892. doi: 10.1039/d1tc00592h

  11. [11]

      H. Uoyama, K. Goushi, K. Shizu, H. Nomura, C. Adachi, Nature 492 (2012) 234–238. doi: 10.1038/nature11687

  12. [12]

      Y. Im, M. Kim, Y.J. Cho, et al., Chem. Mater. 29 (2017) 1946–1963. doi: 10.1021/acs.chemmater.6b05324

  13. [13]

      F. Fang, L. Zhu, M. Li, et al., Adv. Sci. 8 (2021) 2102970. doi: 10.1002/advs.202102970

  14. [14]

      Y. Mei, D. Liu, J. Li, J. Wang, J. Mater. Chem. C 10 (2022) 16524–16535. doi: 10.1039/d2tc03448d

  15. [15]

      X. Dong, S. Shen, Y. Qin, et al., Chin. Chem. Lett. 34 (2023) 108311. doi: 10.1016/j.cclet.2023.108311

  16. [16]

      J. Liu, Z. Feng, C. Peng, et al., Chin. Chem. Lett. 34 (2023) 107634. doi: 10.1016/j.cclet.2022.06.057

  17. [17]

      L. Yan, N. Su, Y. Yang, et al., SmartMat 5 (2024) e1272. doi: 10.1002/smm2.1272

  18. [18]

      M. Mamada, H. Katagiri, C.Y. Chan, et al., Adv. Funct. Mater. 32 (2022) 2204352. doi: 10.1002/adfm.202204352

  19. [19]

      J. Kim, C.W. Joo, S.J. Woo, et al., J. Mater. Chem. C 10 (2022) 7304–7310. doi: 10.1039/d2tc00669c

  20. [20]

      D. Zhang, M. Cai, Y. Zhang, D. Zhang, L. Duan, Mater. Horiz. 3 (2016) 145–151. doi: 10.1039/C5MH00258C

  21. [21]

      H. Noda, H. Nakanotani, C. Adachi, Sci. Adv. 4 (2018) eaao6910. doi: 10.1126/sciadv.aao6910

  22. [22]

      C.Y. Chan, M. Tanaka, H. Nakanotani, C. Adachi, Nat. Commun. 9 (2018) 5036. doi: 10.1038/s41467-018-07482-6

  23. [23]

      S. Zou, F. Xie, M. Xie, et al., Adv. Sci. 7 (2020) 1902508. doi: 10.1002/advs.201902508

  24. [24]

      F.M. Xie, Z.D. An, M. Xie, et al., J. Mater. Chem. C 8 (2020) 5769–5776. doi: 10.1039/d0tc00718h

  25. [25]

      C. Chan, M. Tanaka, Y. Lee, et al., Nat. Photonics 15 (2021) 203–207. doi: 10.1038/s41566-020-00745-z

  26. [26]

      C. Murawski, K. Leo, M.C. Gather, Adv. Mater. 25 (2013) 6801–6827. doi: 10.1002/adma.201301603

  27. [27]

      M. Hasan, A. Shukla, V. Ahmad, et al., Adv. Funct. Mater. 30 (2020) 2000580. doi: 10.1002/adfm.202000580

  28. [28]

      B. Zee, Y. Li, G.J.A.H. Wetzelaer, P.W.M. Blom, Adv. Opt. Mater. 9 (2021) 2100249. doi: 10.1002/adom.202100249

  29. [29]

      Y. Liu, C. Li, Z. Ren, S. Yan, M.R. Bryce, Nat. Rev. Mater. 3 (2018) 18020. doi: 10.1038/natrevmats.2018.20

  30. [30]

      T. Hosokai, H. Matsuzaki, H. Nakanotani, et al., Sci. Adv. 3 (2017) e1603282. doi: 10.1126/sciadv.1603282

  31. [31]

      X. Dai, L. Liu, Z. Ji, Q. Meng, Y. Zou, Chin. Chem. Lett. 34 (2023) 107239. doi: 10.1016/j.cclet.2022.02.044

  32. [32]

      Y. Zou, Q. Meng, H. Mao, D. Zhu, Org. Electron. 41 (2017) 307–314. doi: 10.1016/j.orgel.2016.11.021

  33. [33]

      J. Chen, Y. Xiao, K. Wang, et al., Angew. Chem. Int. Ed. 60 (2021) 2478–2484. doi: 10.1002/anie.202012070

  34. [34]

      W. Li, M. Li, W. Li, et al., ACS Appl. Mater. Interfaces 13 (2021) 5302–5311. doi: 10.1021/acsami.0c19302

  35. [35]

      L. Gan, Z. Xu, Z. Wang, et al., Adv. Funct. Mater. 29 (2019) 1808088. doi: 10.1002/adfm.201808088

  36. [36]

      M.J. Frisch, G.W. Trucks, H.B. Schlegel, et al., Gaussian 16 Revision A.03, Gaussian Inc., Wallingford, CT, 2016.

  37. [37]

      R.L. Martin, J. Chem. Phys. 118 (2003) 4775–4777. doi: 10.1063/1.1558471

  38. [38]

      Y. Liu, P. Ganesan, M. Chen, et al., Chem. Eng. J. 499 (2024) 156529. doi: 10.1016/j.cej.2024.156529

  39. [39]

      J.U. Kim, I.S. Park, C.Y. Chan, et al., Nat. Commun. 11 (2020) 1765. doi: 10.1038/s41467-020-15558-5

• 

  Figure 1  X-ray single crystal diffraction structures of the two compounds: (left) 3DPh and (right) 2DPh.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Absorption spectra with Gaussian fitting curves of (a) 3DPh and (b) 2DPh. (c) UPS spectra and (d) transient PL decays of 3DPh and 2DPh. PL spectra at 300 K, 80 K, and 80 K with 2 ms delay of (e) 3DPh and (f) 2DPh.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  The structures of the 3DPh and 2DPh compounds (upper, the values in parentheses are crystal data), and their calculated FMO and NTO distributions via Gaussian at B3LYP/6–31G(d) level (lower).

  下载: 全尺寸图片 幻灯片
  

  Figure 4  (a) The molecular structures and (b) energy level alignment of organic semiconductors in this study. (c) PL spectrum of mCP and absorption of the 3DPh and 2DPh emitters.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  EL characteristics of the optimized devices with the two emitters: (a) J–V–L, (b) η_CE–L–η_EQE. (c) EL spectra and (d) CIE coordinates at 1000 cd/m².

  下载: 全尺寸图片 幻灯片

  Table 1.  The experimental (Ex) and calculated (Cal) parameters of the two emitters.

  下载: 导出CSV

  Table 2.  The key properties of the OLED devices with different emitters.

  下载: 导出CSV
•