Yuan-Cheng
Hu
a,
Zong-Liang
Lin
a,
Tzu-Chien
Huang
b,
Jhih-Wei
Lee
b,
Wei-Chih
Wei
a,
Tzu-Yu
Ko
a,
Chun-Yuan
Lo
a,
Deng-Gao
Chen
a,
Pi-Tai
Chou
a,
Wen-Yi
Hung
*b and
Ken-Tsung
Wong
*ac
aDepartment of Chemistry, National Taiwan University, Taipei, 10617, Taiwan. E-mail: kenwong@ntu.edu.tw
bDepartment of Optoelectronics and Materials Technology, National Taiwan Ocean University, Keelung 202, Taiwan. E-mail: wenhung@mail.ntou.edu.tw
cInstitute of Atomic and Molecular Science, Academia Sinica, Taipei, 10617, Taiwan
First published on 19th May 2020
In this study, three triazatruxene-based molecules Tr-Me, Tr-Ph, and Tr-Tol with methyl, phenyl, and p-tolyl pendant groups, respectively, were intermixed with three acceptors 3P-T2T, 3P-T2P, and 3P-Pyr featuring phenylpyrazole peripherals and triazine, pyrimidine, and pyridine cores, respectively, to generate an array of nine blends to probe the feasibility of formation of an exciplex. The observed red-shifted emission compared to those of pristine donor and acceptor films is a good indication of exciplex formation. The exciplex emissions are mainly modulated by the features of N-heteroarene cores of acceptors, while the HOMOs of the donors are less influenced by the substituents. From the screening of these D:A combinations it was concluded that four exciplex-forming blends composed of Tr-Ph and Tr-Tol as donors and 3P-T2T and 3P-T2P as acceptors exhibit promising properties suitable for OLED applications. Among them, a green device employing a Tr-Tol:3P-T2P blend as the emitting layer performed with the best efficiency with a maximum external quantum efficiency (EQEmax) of up to 12.8%. These new exciplex-forming systems were further explored as cohosts for a newly developed fluorescent emitter (DT)2BTh2CN. The Tr-Ph:3P-T2P blend was found to serve as the best host to give a deep-red (EL λmax 671 nm, CIE (0.68, 0.31)) OLED device with an EQEmax of 5.52%. The data extracted from the transient photoluminescent studies shed light on the emission mechanism which is dominated by Förster energy transfer and simultaneously suffers from competing Dexter energy transfer quenching. Our results highlight the feasibility of using triazatruxene as the donor for exciplex formation as well as the tunability of the emission wavelength of exciplex-forming systems by rational structural modifications of both donors and acceptors.
Scheme 1 Molecular structures and synthetic pathways of the triazatruxenes, 3P-T2T, 3P-T2P, 3P-Pyr, and (DT)2BTh2CN. |
The thermal stability of the material plays a crucial role in the fabrication of OLED devices with a vacuum-deposition process, which is analyzed by thermogravimetric analysis (TGA) under nitrogen. The new triazatruxene Tr-Tol exhibits a reliable thermal stability with a decomposition temperature relative to 5% weight loss (Td95) of 369 °C, which is higher than those of Tr-Me or Tr-Ph reported in the literature,28,29 mainly due to the higher molecular weight. The acceptors 3P-T2P and 3P-Pyr, on the other hand, show similar Td95 of 340 and 339 °C, respectively. The thermal stability of 3P-T2T (352 °C)27 is relatively higher than those of 3P-T2P or 3P-Pyr, which could be attributed to the higher molecular planarity due to the reduced steric interactions between peripheral aryl substitutions and the triazine core. The reduction of the nitrogen atom of the central heteroarene core leads 3P-T2P and 3P-Pyr to suffer from adjacent C–H/C–H bond repulsions, consequently giving them slightly twisted molecular conformations as well as decreased thermal stability. Nevertheless, from the aspect of optoelectronic applications, these materials possessed sufficiently high thermal stabilities, making them reliable for vacuum-deposition without the risk of decomposition during the high-temperature evaporation. The morphological stabilities of Tr-Tol, 3P-T2P and 3P-Pyr were examined with a differential scanning calorimeter (DSC). The observed glass transition temperature (Tg) was 64 °C for 3P-T2P and 67 °C for 3P-Pyr, whereas there was no phase transition detected for Tr-Tol up to 350 °C.
Cyclic voltammetry was utilized to probe the electrochemical properties. The redox potentials of Tr-Me, Tr-Ph, Tr-Tol, 3P-T2T, 3P-T2P and 3P-Pyr were respectively measured and recorded with reference to a ferrocene/ferrocenium (Fc/Fc+) redox couple, as shown in Fig. S1 (ESI†). A reversible oxidation potential at 0.27, 0.37 and 0.35 V was observed for Tr-Me, Tr-Ph and Tr-Tol, respectively. Obviously, the electronic properties of triazatruxene can be modulated with the structural feature of the N-substitution, in which the methyl-substituted Tr-Me shows a lower oxidation potential compared to those of the aryl-substituted counterparts Tr-Ph and Tr-Tol. There is no surprise that Tr-Tol exhibits a slightly lower oxidation potential than Tr-Ph due to the electron-richer p-tolyl groups. With reference to the Fc/Fc+ redox couple, the HOMO energy levels were estimated as −5.07 eV (Tr-Me), −5.17 eV (Tr-Ph), and −5.15 eV (Tr-Tol). The respective ionization potentials of Tr-Me, Tr-Ph and Tr-Tol solid films were measured by AC-2 to be −4.96, −5.26 and −5.16 eV, respectively (Fig. S2 in ESI†). The trend observed by AC-2 measurement agrees with the results from CV measurement. In contrast to the triazatruxene-based donors, there was a reversible reduction potential at −1.82 and −2.15 V for the N-heteroarene-cored acceptors 3P-T2T and 3P-T2P, respectively, while an irreversible reduction potential at −2.43 V was observed for 3P-Pyr. Apparently, the number of nitrogen atoms embedded in the N-heteroarene core strongly governs the reduction behavior, in which the most electron-deficient triazine core of 3P-T2T shows a better propensity to be reduced compared to those of 3P-T2P or 3P-Pyr. The LUMO energy levels of 3P-T2T, 3P-T2P and 3P-Pyr were estimated to be −2.98, −2.65 and −2.37 eV, respectively, which are in line with the degree of electron deficiency in their central N-heteroarene core. Interestingly, the difference in the HOMO energy levels between triazatruxene-based donors is relatively small, whereas the shift in LUMO energy level is more apparent in N-heteroarene-cored acceptors. Therefore, we could reasonably speculate that the exciplex energy upon combining triazatruxene donors and N-heteroarene acceptors would be tuned mainly by the acceptor counterpart.
The photophysical behaviors of triazatruxene-based donors were studied by UV-vis absorption and photoluminescence (PL) spectroscopy, as shown in Fig. 1a. The data are summarized in Table 1. In diluted toluene solution at room temperature, the absorption spectra of Tr-Ph and Tr-Tol corresponding to the π–π* transition are similar, with a maximum absorption peak (λmax) centered around 317 nm, which is blue-shifted compared to that (λmax = 324 nm) of Tr-Me. In addition, two weak absorption peaks were observed at 332 nm and 350 nm for Tr-Ph and Tr-Tol, respectively, whereas Tr-Me shows slightly red-shifted absorptions at 337 and 354 nm, which are also ascribed to the π–π* transition in the conjugated π-electron system.28 The PL spectra of Tr-Ph and Tr-Tol show a maximum peak (PL λmax) centered at 391 nm, while the PL λmax of Tr-Me is red-shifted to 396 nm. The slightly red-shifted emission of Tr-Me from Tr-Ph is in line with the results reported in the literature.28,29 The absorption and emission spectra of the vacuum-deposited films of Tr-Me, Tr-Ph and Tr-Tol were acquired, as shown in Fig. S3a (ESI†). Obviously, compared to those of Tr-Ph and Tr-Tol, the absorption spectrum of Tr-Me film exhibits a significant red-shift and broad features, indicating strong intermolecular interactions in the solid state. In addition, all of these triazatruxenes show emissions with vibronic and red-shifted characteristics compared to those observed in dilute toluene solution. Parallel to the trend observed in the absorption spectra, the PL spectrum of Tr-Me shows an emission maximum centered at 414 nm, which is significantly red-shifted compared to those of Tr-Ph and Tr-Tol, implying a much stronger propensity for intermolecular aggregations in Tr-Me neat film than in Tr-Ph or Tr-Tol films due to the smaller methyl groups. The strong aggregations render the determination of the energy of the triplet excited state difficult. Therefore, the PL spectra of Tr-Me, Tr-Ph and Tr-Tol (10 wt%) doped in mCP were investigated, as shown in Fig. 1b. PL spectra without long wavelength tailing were detected, indicating the suppressed intermolecular interaction, as mCP was employed as a solid matrix. However, a slight red-shift in PL λmax (405 nm) relative to 393 nm for both Tr-Ph and Tr-Tol was still observed. Accordingly, the corresponding energies of the singlet state estimated from the emission onset are 3.35, 3.41, 3.41 eV for Tr-Me, Tr-Ph and Tr-Tol, respectively. The phosphorescence spectra of Tr-Me, Tr-Ph and Tr-Tol doped films were acquired at 77 K (Fig. 1b). Therefore, the triplet energy was estimated to be 2.90 eV for all these triazatruxene-based donors by the onset (429 nm) of phosphorescence. Apparently, the N-phenyl or N-p-tolyl rings are twisted with respect to the triazatruxene core, thereby preventing intermolecular π–π interactions in Tr-Ph and Tr-Tol molecules. Whereas, lacking intrinsic structural congestion, Tr-Me molecules form severe self-aggregations that implies a poor tendency for forming a uniform film mixed with acceptor materials.
Fig. 1 (a) UV-vis and PL of Tr-Me, Tr-Ph and Tr-Tol (10−5 M) in toluene, (b) the PL spectra of Tr-Me, Tr-Ph and Tr-Tol (10 wt%) doped in mCP measured at room temperature and 77 K. |
Molecule | λ max sol./film (nm) | PL λmaxa sol./film/doped film (nm) | PLonset nm (eV)d | Phosonset nm (eV)e | E g (sol.) (eV) | HOMO (eV) | LUMO (eV) | IP (eV)f | T d (°C) |
---|---|---|---|---|---|---|---|---|---|
a Measured in toluene at a concentration of about 10−5 M. b Calculated from potential vs. ferrocene/ferrocenium redox couple. c Calculated from the difference between HOMO/LUMO and corresponding optical bandgap. d Estimated from the onset of emission spectra from 10% materials doped in mCP at room temperature. e Estimated from the onset of emission spectra from the films of 10 wt% materials doped in mCP at 77 K. f Measured from AC-2. g Not available. h Reported value. | |||||||||
Tr-Me | 319, 337, 354/325 | 396/414/406 | 371 (3.35) | 429 (2.90) | 3.37 | −5.07b | −1.70c | −4.96 | 327 |
Tr-Ph | 316, 332, 350/319 | 391/397/394 | 364 (3.41) | 429 (2.90) | 3.31 | −5.17b | −1.86c | −5.26 | 363 |
Tr-Tol | 316, 332, 350/319 | 391/394/394 | 364 (3.41) | 429 (2.90) | 3.31 | −5.15b | −1.84c | −5.16 | 369 |
3P-T2T | 267/268 | 397/419/440 | 395 (3.15) | 425 (2.92) | 3.58 | −6.56c | −2.98b | N/Ag | 352h |
3P-T2P | 261/259 | 361/380, 489/434 | 379 (3.28) | 430 (2.89) | 3.50 | −6.15c | −2.65b | N/A | 340 |
3P-Pyr | 258/259 | 367/367/396 | 349 (3.56) | 430 (2.89) | 3.63 | −6.00c | −2.37b | N/A | 339 |
The photophysical behaviors of N-heteroarene-cored acceptors were also studied by UV-vis absorption and PL spectroscopy, as shown in Fig. 2a. The data are summarized in Table 1. The absorption spectra of 3P-T2P and 3P-Pyr show similar results with the absorption peaks at around 259 and 325 nm, which could be attributed to the π–π* transition and n–π* transition,34,35 respectively. The absorption edge can be unambiguously assigned to determine the optical energy gap to be 3.58, 3.50 and 3.63 eV for 3P-T2T, 3P-T2P and 3P-Pyr, respectively. The emission with the PL λmax of 3P-T2P and 3P-Pyr in dilute CH2Cl2 solution is detected at 360 nm, which is much blue-shifted compared to that of 3P-T2T (395 nm) reported previously.27 The absorptions (Fig. S3b, ESI†) of 3P-T2P and 3P-Pyr films are similar with λmax at around 259 nm, which is slightly blue-shifted from the 268 nm of 3P-T2P film.27 Apart from 3P-T2T and 3P-Pyr, which show single emission peaks centered at 420 and 367 nm, respectively, in their emission spectra of neat films, the emission spectrum of 3P-T2P shows a weak emission at 380 nm corresponding to its typical molecular chromophore and a strong and broad emission centered at 490 nm ascribed to the excimer emission. The emission spectra of 3P-T2P, 3P-T2P and 3P-Pyr (10 wt%) blended in mCP were examined, as shown in Fig. 2b. The lower energy emission of 3P-T2P disappeared when it served as a guest dopant dispersed in an mCP host, verifying the possibility of forming excimer emission in pristine film. The respective PL λmax of 3P-T2P, 3P-T2P and 3P-Pyr blended films are distinguishable, indicating the effects of the electron-accepting power of the central N-heteroarenes and molecular rigidity on the singlet energy. The higher electron-withdrawing propensity of the triazine core together with the lack of C–H/C–H steric interactions leads 3P-T2T to have a red-shifted λmax of 395 nm as compared to the 379 nm of 3P-T2P and 349 nm of 3P-Pyr. The triplet energies of 3P-T2T, 3P-T2P and 3P-Pyr were determined to be 2.92, 2.89 and 2.89 eV, respectively, based on the onset energy of the phosphorescence spectra (Fig. 2b) obtained at 77 K.
Fig. 2 (a) UV-vis and PL of 3P-T2T, 3P-T2P and 3P-Pyr (10−5 M) in CH2Cl2, (b) the PL spectra of 3P-T2T, 3P-T2P and 3P-Pyr (10 wt%) doped in mCP measured at room temperature and 77 K. |
To explore the formation of exciplexes using Tr-Me, Tr-Ph and Tr-Tol as donors and 3P-T2T, 3P-T2P and 3P-Pyr as acceptors, an array of 9 vacuum-processed D:A (1:1) blends was prepared. The absorption spectra (Fig. S4, ESI†) of all blended films show strong absorption peaks around 260 and 320 nm, which can be considered to be the linear combination of composing donor and acceptor materials. The corresponding emission spectra are depicted in Fig. 3 and the data are summarized in Table 2. No significant new absorption peaks are found in these blends, indicating the lack of ground state interactions between donors and acceptors. In contrast, the blended films did exhibit distinct emission spectra different from their components. As compared to the individual emission of donor or acceptor, the red-shifted emission of blended film is the signature of forming an exciplex upon photoexcitation. The blends consisting of Tr-Me as donor mixed with acceptors 3P-T2T, 3P-T2P and 3P-Pyr gave emission peaks (Fig. 3a) centered at 551, 557 and 530 nm, respectively. In addition, the PL spectra of Tr-Me:3P-T2T and Tr-Me:3P-T2P blends show the residual emission of Tr-Me, indicating incomplete exciplex formation due to the self-segregation of less sterically hindered Tr-Me together with more coplanar acceptors 3P-T2T and 3P-T2P. However, the emission of individual D or A components disappeared in the Tr-Me:3P-Pyr blend, indicating that the twisted conformation of 3P-Pyr is beneficial for suppressing its self-segregation and also blocking the severe intermolecular interactions of Tr-Me molecules. The Tr-Ph-blended films composed of acceptors 3P-T2T, 3P-T2P and 3P-Pyr show emission (Fig. 3b) centered at 552, 527 and 482 nm, respectively. Finally, the blends formed with Tr-Tol as donor and 3P-T2T, 3P-T2P and 3P-Pyr as respective acceptors give emission peaks (Fig. 3c) centered at 554, 526 and 492 nm, respectively. In general, the Tr-Me-based blends exhibit red-shifted emissions compared to those counterpart blends with Tr-Ph and Tr-Tol as donors due to its higher lying HOMO energy level. Obviously, the energy of exciplex emission, typically correlated with the energy difference between donor (HOMO) and acceptor (LUMO),19 is governed mainly by the structural features of N-heteroarene centers. Time-resolved photoluminescence (TRPL) experiments, except for the incomplete exciplex formation of the blends with Tr-Me as donor, were conducted to probe the relaxation processes of these exciplex emissions (Fig. S5, ESI†). The results are summarized in Table 2. We speculate that the low electron deficiency of 3P-Pyr together with relatively weak donors Tr-Ph and Tr-Tol may not induce sufficient polarization for efficient rISC upon photoexcitation.36 The pump and delayed fluorescence spectra are provided in Fig. S6 (ESI†). The prompt emission acquired by an intensified charge couple detector at zero delay time and a gate width of 100 ns is attributed to fluorescence, while the delayed emission acquired at a delay time of 1 μs and gate width of 10 μs is ascribed to delayed fluorescence. The prompt and delayed emissions were recorded from the emission onsets to calculate the singlet–triplet splitting energy (ΔEST). Obviously, the large ΔEST of Tr-Ph: 3P-Pyr (0.34 eV) and Tr-Tol: 3P-Pyr (0.33 eV) make it hard to undergo an efficient rISC process that leads to low PLQY. To our delight, the TRPL of the 4 blends consisting of Tr-Ph and Tr-Tol as donors and 3P-T2T and 3P-T2P as acceptors showed rather evident delayed components in a microsecond timescale and moderate PLQYs ranging from 33 to 41%, indicating the possibility of producing better device characteristics due to superior TADF characters.
Fig. 3 Emission spectra of (a) Tr-Me:acceptor, (b) Tr-Ph:acceptor and (c) Tr-Tol:acceptor intermixed films. |
Donor | Acceptor | PLQYa (%) | PLmax/onset (nm) | ΔESTb (eV) | TRPLc | |||
---|---|---|---|---|---|---|---|---|
A 1 | τ 1 (μs) | A 2 | τ 2 (ns) | |||||
a The PLQY for the blend were measured with an integrating sphere (Hamamatsu C9920-02). b ΔEST are calculated from the energy difference between the emission onsets of prompt and delayed emissions, as shown in Fig. S6 (ESI). c The TRPL profiles were measured under an ambient atmosphere, and the decay components were fitted with two exponential decay models as I(τ) = A1exp(−t/τ1) + A2exp(−t/τ2), as shown in Fig. S5 (ESI). d Derived optical energy gap (eV) from 1242.4/PLonset (nm). | ||||||||
Tr-Ph | 3P-T2T | 39 | 553/479 (2.59)d | 0.10 | 0.11 | 1.95 | 0.89 | 33.1 |
3P-T2P | 41 | 526/456 (2.73) | 0.18 | 0.09 | 1.77 | 0.91 | 52.0 | |
3P-Pyr | 8 | 480/405 (3.07) | 0.34 | 0.06 | 0.07 | 0.94 | 25.6 | |
Tr-Tol | 3P-T2T | 33 | 553/480 (2.59) | 0.10 | 0.10 | 1.73 | 0.90 | 25.8 |
3P-T2P | 40 | 525/457 (2.73) | 0.10 | 0.09 | 2.39 | 0.91 | 56.4 | |
3P-Pyr | 10 | 491/405 (3.07) | 0.33 | 0.04 | 1.0 | 0.96 | 29.0 |
Fig. 4 (a) Schematic exciplex-based OLED structure, and (b) energy levels of donors, acceptors, and fluorescent emitter used in this study. |
Fig. 5 depicts the current density–voltage–luminance (J–V–L) characteristics, device efficiencies, and EL spectra. The key characteristics are summarized in Table 3. Fig. 4b shows a schematic energy level of the donors and acceptors as well as the fluorescent emitter used in this work. The HOMO levels of Tr-Ph and Tr-Tol are similar (−5.17 and −5.15 eV). However, the LUMO levels of 3P-Pyr (−2.37 eV) > 3P-T2P (−2.65 eV) > 3P-T2T (−2.98 eV) will strongly influence the emissive property of the resultant exciplex. Devices A1 and B1 use the EML where 3P-T2T was selected as the acceptor to blend with Tr-Ph and Tr-Tol, respectively. The electrons are injected from CN-T2T to 3P-T2T without any energy barrier, leading to a low turn-on voltage (Von) of 2.0 eV and a high current density. Device A1 gives a luminance (L) of 84318 cd m−2 at 6.0 V (1309 mA cm−2) with an EL peak at 558 nm and CIE coordinates of (0.43, 0.53) and the maximum EQE, current efficiency (CE), and power efficiencies (PE) are 10.3%, 29.8 cd A−1 and 37.5 lm W−1. The extremely low driving voltage of 2.7 V at 1000 cd m−2 means that the carriers are injected smoothly into the HOMO/LUMO levels of the exciplex EML without any energy barrier. Device A1 retains an EQE of 9.9% at a high luminance of 1000 cd m−2. In addition, device A1 exhibits better efficiency than device B1 that exhibits EQE, CE, and PE of 9.7%, 27.3 cd A−1, and 35.7 lm W−1 and CIE coordinates of (0.44, 0.53). This result is consistent with the PLQYs measured for the thin films of Tr-Ph:3P-T2T (39%) and Tr-Tol:3P-T2T (33%) blends. Obviously, similar to the previously reported results regarding 3P-T2T39–41 as a good acceptor with other donors for efficient exciplex systems, the triazatruxene-based molecules Tr-Ph and Tr-Tol are also good donors, which can be further utilized to examine the new exciplex-forming systems with other acceptors.
EML | V on [V] | L/J at 6 V [cd m−2/mA cm−2] | EQE/CE/PEmax [%/cd A−1/lm W−1] | At 1000c cd m−2 [%/cd A−1/lm W−1 V−1] | λ max [nm] | CIE [x, y] | |
---|---|---|---|---|---|---|---|
a The notation A and B indicate the donors Tr-Ph and Tr-Tol, respectively. The notation 1–3 indicate the acceptors 3P-T2T, 3P-T2P and 3P-Pyr, respectively. The notation r indicates 10 wt% (DT)2BTh2CN as dopant. b Turn-on voltage at which emission became detectable. c The values of EQE, CE, PE and driving voltages of device at 1000 cd m−2. | |||||||
A1 | Tr-Ph:3P-T2T | 2.0 | 84318/1309 | 10.3/29.8/37.5 | 9.92/28.7/33.4/2.7 | 558 | 0.43, 0.53 |
A2 | Tr-Ph:3P-T2P | 2.2 | 37975/601 | 10.4/30.8/37.2 | 7.67/22.7/21.6/3.3 | 522 | 0.33, 0.54 |
A3 | Tr-Ph:3P-Pyr | 2.8 | 463/64 | 0.54/0.91/0.86 | 0.40/0.67/0.31/6.7 | 469 | 0.19, 0.24 |
B1 | Tr-Tol:3P-T2T | 2.0 | 75082/1157 | 9.7/27.3/35.7 | 9.5/26.6/32.2/2.6 | 566 | 0.44, 0.53 |
B2 | Tr-Tol:3P-T2P | 2.2 | 26280/360 | 12.8/38.3/46.3 | 8.9/26.6/24.6/3.4 | 526 | 0.35, 0.54 |
B3 | Tr-Tol:3P-Pyr | 2.6 | 2008/164 | 0.55/0.97/0.98 | 0.42/0.74/0.44/5.3 | 467 | 0.19, 0.25 |
10 wt% (DT)2BTh2CN doped in cohost | |||||||
A1r | Tr-Ph:3P-T2T | 1.8 | 3165/1961 | 4.64/1.24/1.96 | 1.07/0.28/0.20/4.5 | 674 | 0.69, 0.31 |
A2r | Tr-Ph:3P-T2P | 2.3 | 966/410 | 5.52/1.66/2.09 | 0.75/0.22/0.11/6.1 | 671 | 0.68, 0.31 |
B1r | Tr-Tol:3P-T2T | 1.8 | 3666/3078 | 3.48/0.98/1.47 | 0.68/0.19/0.13/4.5 | 669 | 0.69, 0.31 |
B2r | Tr-Tol:3P-T2P | 2.3 | 782/276 | 4.42/1.72/2.16 | 0.64/0.25/0.12/6.4 | 667 | 0.67, 0.32 |
Devices A2 and B2 using the exciplex-based EML composed of 3P-T2P as acceptor and Tr-Ph and Tr-Tol as donors, respectively, exhibit green emissions, which are blue-shifted compared to those of the devices employing 3P-T2T as an acceptor due to the higher LUMO energy of 3P-T2P. The lower current density and higher Von for 3P-T2P-based devices are due to the energy barrier (0.13 eV) for the electron injection from CN-T2T to 3P-T2P. Device B2 based on the Tr-Tol:3P-T2P blend as the EML shows the best performance among the devices reported in this work, giving a Von of 2.2 V and maximum efficiencies as high as 12.8% (EQE), 38.3 cd A−1 (CE), and 46.3 lm W−1 (PE) with CIE coordinates of (0.35, 0.54). Correspondingly, the device with the Tr-Ph:3P-T2P blend shows lower maximum efficiencies (10.4%, 30.8 cd A−1, 37.2 lm W−1) with CIE coordinates of (0.33, 0.54). The superior efficiency obtained in devices A2 and B2 agrees with the observed better PLQYs of exciplex-forming systems with 3P-T2P as the acceptor (Table 2). Devices A3 and B3 use 3P-Pyr as an acceptor and Tr-Ph and Tr-Tol as a donor, respectively, as the EML for making exciplexes. The higher LUMO level (−2.37 eV) of 3P-Pyr leads the exciplex to have blue emission with CIE coordinates of (0.19, 0.24–0.25). The low J–V performance and high Von are because of the larger energy barrier (0.41 eV, see Fig. 4a and b) between CN-T2T and 3P-Pyr that impedes the electron injection to the EML. Together with the low fluorescence quantum yield of 8% for Tr-Ph:3P-Pyr and 10% for Tr-Tol:3P-Pyr, these give the low EQE (<1%) of 3P-Pyr-based blue devices.
The highly efficient exciplex system can be employed as the cohost for improving the efficiencies of OLEDs42 because the triplet excitons of the exciplex can be up-converted back to singlets by way of the efficient rISC process. Then, the singlet excitons of the exciplex cohost can transfer to the singlet state of the fluorescent dopant via a Förster energy transfer (FRET) pathway, making high-efficiency fluorescent OLEDs accessible.11 To realize an effective FRET, a good spectral overlap between the emission of the host and the absorption of the dopant is required. Fig. 6(a) displays the good overlaps between the absorption spectrum of the newly developed fluorescent dopant (DT)2BTh2CN (Scheme 1) and the PL spectra of exciplex cohosts (Tr-Ph/Tr-Tol:3P-T2T and Tr-Ph/Tr-Tol:3P-T2P), ensuring that efficient energy transfer from the exciplex to the dopant is feasible.
The exciplex-forming systems were further exploited as cohosts for a red fluorescent dopant (DT)2BTh2CN. The photophysical and electrochemical properties of (DT)2BTh2CN are depicted in Fig. S8 (ESI†), and the data together with the thermal stability are summarized Table S1 (ESI†). First, the devices employing the best exciplex Tr-Tol:3P-T2P blend as the cohost incorporated with various doping concentrations of (DT)2BTh2CN were fabricated to optimize the EL performance (Fig. S9, ESI†). The device doped with 1 wt% (DT)2BTh2CN exhibits maximum EQE and PE of 5.5% and 8.4 lm W−1, respectively. However, at such a low doping concentration, the EL spectrum shows residual emission from the exciplex cohost with an EL peak at 520 nm.43 The device doped with 5 wt% (DT)2BTh2CN displays better energy transfer; however, a weak emission from the exciplex cohost still remains. This device shows the EL emission peak centered at 660 nm and CIE coordinates of (0.64, 0.35), but a lower maximum EQE (4.39%) compared to the 1 wt% doped device. When the doping concentration increases to 10 wt%, the device retains a comparable EQE (4.42%) and exhibits EL solely from the dopant but with the peak further red-shifted to 672 nm and CIE coordinates of (0.67, 0.32), indicating complete energy transfer. The color-shifting is because of the intermolecular interactions and strong local polarization field.44 The reduced device efficiency found in the devices with higher doping concentrations of (DT)2BTh2CN can be plausibly ascribed to the triplet–triplet Dexter energy transfer that leads to direct exciton quenching by the non-emissive triplets of the fluorescent dopant. To balance the efficient Förster energy transfer for giving pure emission of (DT)2BTh2CN and the competitive quenching process by the inevitable Dexter energy transfer, 10 wt% (DT)2BTh2CN was selected as the doping concentration for the exciplex cohosts (Tr-Ph/Tr-Tol:3P-T2T and Tr-Ph/Tr-Tol:3P-T2P) to investigate the performance of the deep-red fluorescent devices. Fig. 6(b–d) depict the EL characteristics of these deep-red fluorescent devices, and the key data are summarized in Table 3. All 10 wt% (DT)2BTh2CN-doped devices exhibit pure emission from the dopant with the EL peak centered at ca. 670 nm. Among them, the Tr-Ph:3P-T2P-hosted device A2r shows the best performance, achieving maximum EQE of 5.52%, CE of 1.66 cd A−1 and PE of 2.09 lm W−1 with CIE coordinates of (0.68, 0.31). The 5.52% EQE based on a Tr-Ph:3P-T2T cohost is higher than the theoretical limit of 5% typically anticipated for a conventional fluorescence emitter. The device performances are in good agreement with the PLQYs of the (DT)2BTh2CN emission extracted from the exciplex-forming cohosts as in the data summarized in Table S2 (ESI†). TRPL measurements were further performed on the (DT)2BTh2CN-doped exciplex cohosts (Tr-Ph/Tr-Tol:3P-T2T and Tr-Ph/Tr-Tol:3P-T2P) to probe the working mechanism. The data are summarized in Table S2 (ESI†). As indicated, the (DT)2BTh2CN-dispersed films show a dominant prompt decay component with a lifetime of around 7 ns together with a delay component with a lifetime of around 0.13 μs. The delay component only counted for about 1% of the whole relaxation process when Tr-Ph was applied as the donor in the blended film. In the case of Tr-Tol utilized as the donor for exciplex-forming systems, the delay components are as low as only 0.2% of the overall relaxation process. Accordingly, the EL from the exciplex-forming cohosts doped with 10 wt% (DT)2BTh2CN is mainly contributed by the Förster energy transfer, but simultaneously suffers from the inferior triplet-harvesting of the exciplex due to the competing Dexter energy transfer quenching in the presence of high dopant concentration.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/d0qm00188k |
This journal is © the Partner Organisations 2020 |