Introduction
Quantum efficiency has been regarded as the most important criterion for the evaluation of organic light-emitting diodes (OLEDs) [1–4]. In comparison to the conventional fluorescent devices, the discovery of phosphorescent OLEDs is a major breakthrough. Phosphorescent materials provide an inherent approach for ~100% excitons harnessing ability with the help of strengthened spin-orbit coupling [5–7]. Nowadays, OLED and LED could complement each other with their own advantages and meet different commercial requirements [8–10]. Throughout the past decades, numerous phosphorescent OLEDs which emit light ranging from blue to red have already been discovered [11–18]. However, there are few reports on saturated deep-red ones possessing Commission International de L'Eclairage_x (CIE x)≥0.67 with high efficiency and low efficiency roll-off simultaneously [19,20]. Such a predicament could be mainly ascribed to the fact that the non-radiative rate constant of emitter molecules tends to increase as the energy gap decrease, which therefore deteriorates the device efficiency [21,22]. Most research efforts are devoted to new phosphorescent emitter molecules design [23,24]. However, a well matched host matrix also plays a crucial role in high-performance OLEDs. First, as is well known, a so-called host-guest system by homogeneously dispersing a phosphorescent emitter into a host matrix at a proper concentration is commonly adopted in most phosphorescent OLEDs [25–27]. It is believed that this structure contributes to suppressing the undesired excitons decay pathways such as aggregation quenching, triplet-triplet annihilation (TTA), etc. Moreover, for the purpose of suppressing a backward endothermic energy transfer and achieving an efficient host→guest energy transfer, a proper triplet energy levels (ET) of host material is also necessary. At last, the host matrix could be seriously affects the charge-balance within the emissive layers (EMLs). Bipolar host materials have gained immense interests in recent years due to their balanced charge transport abilities.
In deep red phosphorescent OLEDs, the host matrix always possesses a much larger energy gap than the guest materials, and the latter usually constitute charge-trapping sites within EMLs. The forms of the charge-trapping sites are different in various host-guest compatibilities. The influence of the charge-trapping effect on the device quantum efficiency is unnegligible [28–33]. In our previous work, we constructed an asymmetrical charge-trapping based on two carbazole/thiadiazole hybrids [34]. The relationship between electron-trapping and device performances was investigated and we confirmed that the hole-trapping was unfavorable in all devices. However, the difference between a symmetric and an asymmetrical charge-trapping effect is still unclear. A further investigation is necessary.
In this paper, we designed and synthesized three bipolar host materials, 2,2'-(1,2,4-thiadiazole-3,5-diyl)bis(N,N-diphenylaniline) o-TPATHZ, 3,3′-(1,2,4-thiadiazole-3,5-diyl)bis(N,N-diphenylaniline) m-TPATHZ, 4,4'-(1,2,4-thiadiazole-3,5-diyl)bis(N,N-diphenylaniline) and p-TPATHZ, by fine-tuning the linkage mode between 1,2,4-thiadiazole and triphenylamine. Phosphorescent emitters tris(1-phenylisoqiunoline)iridium (Ir(piq)3) and [bis(2-methyldibenzo-[f,h]-quinoxaline)Ir(III)(acetylacetonate)] (Ir(MDQ)2(acac)) were introduced for fabricating efficient deep red OLEDs. Due to the different energy levels and energy gaps of the materials, varied charge-trapping characteristics were obtained within different EMLs. In particular, improved performance is achieved by symmetric charge-trapping effect. The maximum external quantum efficiency (EQE) in Ir(piq)3-based device utilizing p-TPATHZ as host matrix was 16.2%. Corresponding to a maximum current efficiency and power efficiency of 13.3 cd/A and 13.6 lm/W, respectively. And the EQE remained as high as 15.7% at the high luminance of 1000 cd/m2, with an extremely low efficiency roll-off of about 3.1%. The Commission Internationale de l’Eclairage (CIE) coordinates were (0.68, 0.32).
Experimental section
General procedures
The immediate compounds and target materials were confirmed by 1H NMR or 13C NMR spectra on a Bruker-AF301 AT 400 MHz spectrometer. Mass spectra were carried out on an Agilent (1100 LC/MSD Trap) using atmospheric pressure chemical ionization (APCI). The Fourier transform infrared (FTIR) spectra were recorded on a VERTEX 70 FTIR Spectrometer. Elemental analyses of carbon, hydrogen and nitrogen were performed on an Elementar (Vario Micro cube) analyzer. Differential scanning calorimetry (DSC) was measured under nitrogen on a PE Instruments DSC 2920 unit at a heating rate of 10°C/min from 30°C to 300°C, and the glass transition temperature (Tg) was determined from the second heating scan. Thermogravimetric analysis (TGA) was undertaken with a PerkinElmer Instruments (Pyris1 TGA). The thermal stability of the samples were determined by measuring their weight loss while heating at a rate of 10°C/min from 30°C to 700°C under a nitrogen atmosphere. The UV-Vis absorption spectra were recorded on a Shimadzu UV-VIS-NIR Spectrophotometer (UV-3600) and the photoluminescence (PL) spectra were recorded on an Edinburgh instruments (FLSP920 spectrometer). Cyclic voltammetry (CV) measurements were carried out on a computer-controlled EG&G Potentiostat/Galvanostat model 283 at room temperature with a conventional three-electrode system, which contained a Pt carbon working electrode of 2 mm in diameter, a platinum wire counter electrode and an Ag/AgNO3 (0.1 mol) reference electrode. Oxidations CVs of all compounds were performed in distilled dichloromethane with 0.1 mol tetrabutylammoniumhexafluorophosphate (Bu4NPF6) as the supporting electrolyte. All solutions were purged with a nitrogen stream for 10 min before measurement.
Device fabrication and measurement
The material MoO3, 4,4′-(cyclohexane-1,1-diyl)bis(N,N-di-p-tolylani-line) (TAPC), N,N,N-Tris(4-(9-carbazolyl)phenyl)amine (TCTA), 3,3′-(5′-(3-(pyridine-3-yl)phenyl)-[1,1':3′,1″-terphenyl]-3,3″-diyl) dipyridine (TPBi), LiF as well as two emitters Ir(piq)3 and Ir(MDQ)2(acac) were commercially available. All the devices were fabricated upon the surface of indium-tin-oxide (ITO)-coated glass substrates, the sheet resistance of which was approximately 25 W·square-1. In fabrication procedure, after sequential chemical cleaning and drying, the substrates were exposed to oxygen plasma for 5 min. Then they were loaded into a high vacuum (under a base pressure ca. 5 × 10 -6 Torr) thermal evaporation chamber, followed by successive deposition at a rate of 0.9−1.1 Å/s. In addition, both host and guest materials were co-evaporated and the deposition rates were precisely controlled by individual quartz-crystal thickness monitors. The light-emitting areas were determined by the overlap of two electrodes as 0.09 cm2 and all electroluminescence (EL) properties were measured by combining the Keithley 2400 Digital Source meter with the PR655 under ambient conditions.
All the reagents and solvents used for the syntheses and measurements were purchased from commercial suppliers without further purification unless other stated. The chemical structures of the compounds were confirmed by 1H NMR or 13C NMR spectra, elemental analysis and mass spectrometry.
Synthesis
2,2'-(1,2,4-thiadiazole-3,5-diyl)bis(N,N-diphenylaniline) (o-TPATHZ): A mixture of (2-(diphenylamino)phenyl)boronic acid (1.6 g, 5.6 mmol), 3-bromo-5-chloro-1,2,4-thiadiazole (0.5 g, 2.5 mmol), Pd(PPh3)4 (0.19 g,0.17 mmol), K2CO3 (2.0 mol aqueous solution, 20 mL), toluene (40 mL) and ethanol (20 mL) were stirred under nitrogen at 100°C for 12 h. After cooling to room temperature, it was extracted with CH2Cl2 and water. Then, the organic layer was dried over anhydrous MgSO4 and the solvent was removed by vacuum distillation. The residue was further purified by column chromatography using CH2Cl2 and petroleum ether as the eluent to afford a white powder 1.28 g (yield: 89%). 1H NMR (400 MHz, CDCl3) d[ppm]: 8.30-8.28 (d, J= 7.60 Hz, 1H), 7.76-7.74 (d, J= 7.6 Hz, 1H), 7.50-7.46 (t, J= 8.0 Hz, 1H), 7.43-7.39 (t, J= 8.0 Hz, 1H), 7.36-7.32 (t, J= 7.2 Hz, 1H), 7.29-7.15 (m, 8 H), 7.02-6.92 (m, 13H), 6.77-6.74 (t, J= 6.8 Hz, 2H); 13C NMR (100 MHz, CDCl3) d[ppm]: 180.62, 169.99, 147.64, 146.67, 146.25, 145.98, 132.56, 131.68, 131.60, 131.19, 130.78, 130.40, 129.40, 129.37, 129.03, 128.57, 127.25, 124.97, 123.14, 122.58, 122.14, 121.47; MS (APCI): calcd for C38H28N4S: 572.2035; found: 572.2099 [M+ H] +; Elemental analysis (%): calcd for C38H28N4S: C 79.69, H 4.93, N 9.78, S 5.60; found: C 80.01, H 4.87, N 9.78, S 5.34.
3,3′-(1,2,4-thiadiazole-3,5-diyl)bis(N,N-diphenylaniline) (m-TPATHZ): The synthesize method of m-TPATHZ was referred the upper procedure with a high yield of 84%. 1H NMR (400 MHz, CDCl3) d[ppm]: 8.06 (s, 1H), 7.96-7.94 (d, J= 7.6 Hz, 1H), 7.68 (s, 1H), 7.56-7.55 (d, J= 7.6 Hz, 1H), 7.34-7.22 (m, 10 H), 7.19-6.99 (m, 14 H); 13C NMR (100 MHz, CDCl3) d[ppm]: 187.95, 173.47, 148.82, 148.27, 147.67, 147.21, 134.02, 131.67, 130.10, 129.59, 129.47, 129.27, 126.60, 126.20, 124.59, 124.21, 123.75, 123.51, 122.90, 122.86, 121.65, 121.23; MS (APCI): calcd for C38H28N4S: 572.2035; found: 572.2087 [M+ H] + ; Elemental analysis (%): calcd for C38H28N4S: C 79.69, H 4.93, N 9.78, S 5.60; found: C 80.10, H 4.87, N 9.73, S 5.30.
Synthesis of 4,4'-(1,2,4-thiadiazole-3,5-diyl)bis(N,N-diphenylaniline) (p-TPATHZ). The compound p-TPATHZ was synthesized according to the same procedure as the o-TPATHZ. Yield: 80.0%. 1H NMR (400 MHz, CDCl3) d[ppm]: 8.21-8.18 (d, J= 8.8 Hz, 2H), 7.85-7.83 (d, J= 8.4 Hz, 2H), 7.33-7.24 (m, 8H), 7.16-7.04 (m, 16H); 13C NMR (100 MHz, CDCl3) d[ppm]: 187.21, 173.26, 151.12, 149.65, 147.26, 146.66, 129.58, 129.37, 129.32, 128.56, 126.60, 125.62, 125.07, 124.36, 123.55, 123.42, 122.26, 121.28; MS (APCI): calcd for C38H28N4S: 572.2035; found: 572.2106 [M+ H] +; Elemental analysis (%): calcd for C38H28N4S: C 79.69, H 4.93, N 9.78, S 5.60; found: C 80.25, H 4.86, N 9.78, S 5.11.
Results and discussion
Synthesis
The synthetic routes and chemical structures of the three target compounds are depicted in Scheme 1. All the compounds were prepared through the palladium-catalyzed Suzuki coupling reaction in high yield [35]. The three boric acid intermediate compounds were obtained by purchasing. Additionally, another important intermediate 3-bromo-5-chloro-1,2,4-thiadiazole was prepared according to our previous work [36]. The target compounds were all confirmed by 1H NMR, 13C NMR, elemental analysis, and mass spectrometry. These three host materials were further purified by repeated temperature-gradient vacuum sublimation.
Thermal properties
The thermal stabilities of all hosts were confirmed by the DSC and TGA under nitrogen atmosphere. DSC and TGA thermograms of these compounds were shown in Fig. 1, and the detailed data were summarized in Table 1. The relatively high Tg and Td of these hosts manifest that they could form morphologically stable films and suppress the potential for phase separation during long-time device operation [37].
Photophysical properties
As shown in Figs. 2, 3 and Table 1, basic photophysical characteristics of o-TPATHZ, m-TPATHZ, and p-TPATHZ were analyzed by PL measurements. The larger conjugation of molecules can result lower triplet energy level (ET). The conjugation length was in the order of o-TPATHZ<m-TPATHZ<p-TPATHZ. The analogous rules accord with the previous report [38]. And then the decreasing ETs of 2.68 eV for o-TPATHZ, 2.64 eV for m-TPATHZ and 2.41 eV for p-TPATHZ were obtained according to the highest-energy vibronic sub-band of the phosphorescence spectra. All of compounds possess a high ET of more than 2.4 eV. It is sufficient to suppress a backward energy transfer from red phosphorescent emitters to host molecules. Noteworthy, the difference between their ETs is more than 0.2 eV, it provides a platform to understand the influence of ET on triplet-triplet energy transfer efficiency.
Electrochemical properties
To investigate the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) energy levels of these hosts, oxidation scans of cyclic voltammograms were recorded as displayed in Fig. 4 and the specific data were also listed in Table 1. The HOMO levels were in the range of 5.36 to 5.41 eV, determined from the onsets of the oxidation potentials by comparing with ferrocene (Fc). Meanwhile, their LUMO levels were varied from 2.40 to 2.48 eV, determined from their band gaps (Eg) estimated from the absorption spectra. In general, all three analogs show similar HOMO and LUMO levels.
Quantum chemical calculations
Calculated geometrical properties were investigated to figure out the influence of different substitution positions of triphenylamine (TPA) to the 1,2,4-thiadiazole (THZ) skeleton. Figure 5 shows their optimized molecular structures and orbital distribution. The HOMOs and LUMOs of o-TPATHZ and m-TPATHZ are mainly localized on the electron-rich triphenylamine (TPA) moiety and electron-withdrawing 1,2,4-thiadiazole (THZ) group, respectively. Obviously, the HOMO and LUMO are almost completely spatially separated for p-TPATHZ with minor electron cloud on the 1,2,4-thiadiazole (THZ) group. This separated and outside electron cloud distribution of p-TPATHZ can provide hole- and electron-transporting channels for intermolecular carriers hoping via their respective transporting pathways. It is in favor of improving the device performances.
Fig.6 Calculated HOMO/LUMO distributions and energy levels of the host materials. Computational Details: The geometrical and electronic properties were computed using the Gaussian 09 program package. Molecular orbitals were visualized using Gaussian view. The calculation was optimized by means of the B3LYP (Becke three parameters hybrid functional with Lee-Yang-Perdew correlation functional) [39] with the 6-31G (d) atomic basis set. Then the electronic structures were calculated at t-HCTHhyb/6–311+ + G (d, p) level [40]. Molecular orbitals were visualized using Gaussian view |
Tab.1 Physical properties of o-TPATHZ, m-TPATHZ, and p-TPATHZ |
| compounds | Absa /nm | PLa /nm | Td/Tgb /°C | Egc /eV | ETd /eV | Eoxe /V | (HOMO/ LUMO)/eVe | (HOMO/ LUMO)/eVf |
|---|---|---|---|---|---|---|---|---|
| o-TPATHZ | 374 | 503 | 335/76 | 2.95 | 2.68 | 0.71 | –5.36/2.41 | –1.69/–4.86 |
| m-TPATHZ | 376 | 512 | 365/69 | 2.93 | 2.64 | 0.75 | –5.41/–2.48 | –1.78/–4.90 |
| p-TPATHZ | 370 | 490 | 402/88 | 2.97 | 2.41 | 0.71 | –5.37/–2.40 | –1.63/–4.86 |
Notes: a: Measured in CH2Cl2 at room temperature; b: Td: decomposition temperatures. Tg: glass transition temperatures; c: calculated from the onset of the absorption spectra in CH2Cl2 solution; d: measured in 2-MeTHF glass matrix at 77 K; e: estimated from the CV; f: calculated through the density function theory (DFT) |
Electroluminescence properties
To assess the EL performances of the novel bipolar hosts based deep red phosphorescent OLEDs, we introduced Ir(MDQ)2(acac) and Ir(piq)3 as dopants. We fabricated series of devices with a typical configuration of ITO/MoO3 (10 nm)/TAPC (40 nm)/TCTA (5 nm)/Host: Ir(MDQ)2(acac) or Ir(piq)3 (3wt %, 20 nm)/TPBi (40 nm)/LiF (1 nm)/Al (100 nm) by sequential vapor deposition (Device fabrication and measurement shown in ESI). Note that the EMLs utilize Ir(MDQ)2(acac) as the emitter were named as 1 and that Ir(piq)3 were named as 2. Then, o-TPATHZ-hosted devices were A1 and A2, m-TPATHZ-hosted devices were B1 and B2, and p-TPATHZ-hosted devices were C1 and C2, respectively. The HOMO/LUMO energy levels of the materials were illustrated in Fig. 6. Except for the host matrixes, the energy levels of other materials were from references [41,42].
The EQE-brightness curves were shown in Fig. 7, and the EL properties were summarized in Table 2. Although, the host-guest compatibilities are different between the two series, it is apparently that the p-TPATHZ-hosted devices achieved the highest EQE. For example, the maximum EQE of devices A1, B1 and C1 were 15.2%, 13.8% and 17.4%, respectively. Herein, we attributed this characteristic to their different energy transfer efficiencies, especially those governed by the Dexter mechanism process from host triplet states (T1H) to Ir(MDQ)2(acac) guest triplet states (T1G). In other words, on the basis of effectively suppressing a backward energy transfer, a relatively narrow gap between the ET of host and that of guest could promote the T1H→T1G energy transfer.
Tab.2 EL data of devices A1–C1 and A2–C2 |
| device | host | guest | Von /V | [hc ]a /(cd·A−1) | [hp]b /(lm·W−1) | [hEQE]c /% | roll-offd /% | CIE [x, y]e |
|---|---|---|---|---|---|---|---|---|
| A1 | o-TPATHZ | Ir(MDQ)2(acac) | 3.2 | 28.0 | 26.2 | 15.2/12.3/9.5 | 19.1/37.5 | (0.60, 0.40) |
| B1 | m-TPATHZ | Ir(MDQ)2(acac) | 3.6 | 24.3 | 19.7 | 13.8/12.8/10.0 | 7.2/27.5 | (0.59, 0.39) |
| C1 | p-TPATHZ | Ir(MDQ)2(acac) | 3.5 | 25.8 | 24.0 | 17.4/13.0/10.3 | 25.3/40.8 | (0.61, 0.38) |
| A2 | o-TPATHZ | Ir(piq)3 | 3.5 | 11.1 | 10.8 | 14.4/14.2/12.8 | 1.4/11.1 | (0.68, 0.32) |
| B2 | m-TPATHZ | Ir(piq)3 | 3.1 | 9.4 | 9.3 | 11.6/11.5/10.4 | 0.9/10.3 | (0.67, 0.33) |
| C2 | p-TPATHZ | Ir(piq)3 | 2.9 | 13.3 | 13.6 | 16.2/16.2/15.7 | 0/3.1 | (0.68, 0.32) |
Notes: Von: turn-on voltage; a: the maximum current efficiency; b: the maximum power efficiency; c: order of the external efficiency: maximum, at 100 and 1000 cd/m2; d: the efficiency roll-off at 100 and 1000 cd/m2; e: measured at 20 mA/cm2 |
The EQE roll-offs of the six devices were also shown in Table 2. Unlike the maximum EQE performance achieved by the two series, the efficiency roll-offs characteristic at high brightness between Ir(MDQ)2(acac)-based and Ir(piq)3-based devices were extremely different. As shown in Fig. 7, the roll-offs in in series 1 were much significant than those in series 2. For example, the EQE roll-off in device C1 was about 25.3% at the luminance of 100 cd/m2, 40.8% at the luminance of 1000 cd/m2 and as large as 71.8% at the luminance of 10000 cd/m2, respectively. As a comparison, the EQE roll-off in device C2 was closed to zero at the luminance of 100 cd/m2, about 3.1% at the luminance of 1000 cd/m2 and merely 21.6% even at the luminance of 10000 cd/m2, respectively. Furthermore, along with the sharply decreased EQE in device C1, a grown emission from p-TPATHZ could also be observed in its EL spectra shown in Fig. 8(a), but it’s obviously vanished in Fig. 8(b). We would have a discussion about it to reveal the inner mechanism.
According to the device structure and energy level diagram shown in Fig. 6, the LUMO of Ir(MDQ)2(acac) is 3.3 eV and the LUMO of Ir(piq)3 is 3.1 eV. It was obvious that the electron injection from TPBi into guest molecules was an exothermic process and the LUMO difference between the host and the guest within the EML was about 0.7 eV or even more. It means that the both two phosphorescent emitters functioned as deep trap sites for electrons in all matrixes. In comparison, the hole-trapping effects in different devices were totally different. The HOMO of Ir(piq)3 is 5.1 eV and the HOMO of Ir(MDQ)2(acac) is 5.4 eV. The HOMO differences between the host materials and Ir(piq)3 within the EML were about 0.3 eV while those for Ir(MDQ)2(acac) is nearly zero. Due to the energy level characteristic, we assumed that the Ir(piq)3 also functioned as deep hole-trapping sites in all host matrixes. That is, a symmetric and deep charge-trapping structure was constructed in Ir(piq)3-based devices and stable and balance charges were maintained within the EML even though at high current density. On the contrary, an asymmetrical charge-trapping structure was constructed in Ir(MDQ)2(acac)-based devices and the holes captured on Ir(MDQ)2(acac) molecules can be easily de-trapped. This asymmetrical charge-trapping mode deteriorated the charge balance especially at high current density and was responsible for the significant efficiency roll-offs in devices A1, B1 and C1.
To further prove it, both hole-only devices (HODs) and electron-only devices (EODs) were introduced to estimate the charge-trapping effects in device C1 and C2. The HODs were fabricated in the structures of ITO/MoO3 (8 nm)/TAPC (60 nm)/TCTA (5 nm)/p-TPATHZ: Ir(MDQ)2(acac) or Ir(piq)3 (1 wt %, 3 wt % and 6 wt %, respectively. 20 nm)/MoO3 (8 nm)/Al (100 nm) and the current density-voltage (J-V) curves were shown in Fig. 9. Noteworthy, the hole-current affected by the emitter obviously in Ir(piq)3-based devices but almost no change in Ir(MDQ)2(acac)-based ones. Meanwhile, the EODs were fabricated in the structures of ITO/LiF (1 nm)/p-TPATHZ: Ir(MDQ)2(acac) or Ir(piq)3 (1 wt %, 3 wt % and 6 wt %, respectively. 20 nm)/TPBi (40 nm)/LiF (1 nm)/Al (100 nm). As shown in Fig. 10, deep electron-trapping effect occurred in both of Ir(MDQ)2(acac) and Ir(piq)3. Those observations were in accordance with the above derivations. Table 3 was given to compare the roll-offs between our works and the reported devices with Ir(piq)3. Better efficiency roll-off characteristics were achieved by utilizing our novel host materials.
Tab.3 Compare the efficiency roll-offs with reported Ir(piq)3-based OLEDs |
| device | host | guest | [hEQE ]a/% | [hEQE]b/% | roll-offc/% |
|---|---|---|---|---|---|
| Ref. [43] | TCPB | Ir(piq)3 | 18.6 | 10.2/7.93 | 45.2/57.4 |
| Ref. [43] | TCPY | Ir(piq)3 | 18.4 | 15.5/13.8 | 15.8/25 |
| Ref. [43] | TCPM | Ir(piq)3 | 18.2 | 17.8/15.4 | 2.2/15.4 |
| Ref. [44] | 1d | Ir(piq)3 | 18.2 | 13.6/8.36 | 25.3/54.1 |
| Ref. [44] | 2d | Ir(piq)3 | 15.8 | 12.6/9.9 | 20.3/37.3 |
| Ref. [44] | 3d | Ir(piq)3 | 15.7 | 13.2/11.6 | 15.9/26.1 |
| A2 | o-TPATHZ | Ir(piq)3 | 14.4 | 14.2/12.8 | 1.4/11.1 |
| B2 | m-TPATHZ | Ir(piq)3 | 11.6 | 11.5/10.4 | 0.9/10.3 |
| C2 | p-TPATHZ | Ir(piq)3 | 16.2 | 16.2/15.7 | 0/3.1 |
Notes: a: the maximum EQE; b: order of the external efficiency: at 100 and 1000 cd/m2; c: the efficiency roll-off at 100 and 1000 cd/m2; d: synthesized novel materials reported in Ref. [44] |
Conclusions
We have designed and synthesized three bipolar host materials o-TPATHZ, m-TPATHZ, and p-TPATHZ. And we have introduced phosphorescent emitters Ir(piq)3 and Ir(MDQ)2(acac) to fabricate deep red OLEDs. It was found that p-TPATHZ-hosted devices perform best. Especially, the device incorporating both p-TPATHZ and Ir(piq)3 has achieved a maximum EQE of 16.2% with an extremely low efficiency roll-off of 3.1% at the luminance of 1000 cd/m2. Deep red emission is also achieved with a CIE coordinates of (0.68, 0.32). We have also confirmed that symmetric charge-trapping effect benefit EQE roll-off characteristic through adjusting the host-guest compatibility in different devices. We indeed believed that this work will pave road to high performance deep red OLEDs.