Efficient Deep-Blue Electrofluorescence with an External Quantum Efficiency Beyond 10.

The design of blue fluorescent materials combining both deep-blue emission (CIEy<0.06) and high-efficiency climbing over the typically limited exciton production efficiency of 25% is a challenge for organic light-emitting diodes (OLEDs). In this work, we have synthesized two blue luminogens, trans-9,10-bis(2-butoxyphenyl)anthracene (BBPA) and trans-9,10-bis (2,4-dimethoxyphenyl)anthracene with high photoluminescence quantum yields (PLQYs) of 89.5% and 87.0%, respectively. Intriguingly, we have proposed a strategy to avoid aggregation-caused quenching, which can effectively reduce the undesirable excimeric emission by introducing two host matrices with twisted molecular structure, 9,10-di(naphth-2-yl) anthracene and 10,10'-bis-(4-fluorophenyl)-3,3'-dimethyl-9,9'-bianthracene (MBAn-(4)-F), in the BBPA emission layer. The device containing the EML of BBPA-doped MBAn-(4)-F exhibited a high external quantum efficiency of 10.27% for deep-blue emission with the Commission International de L'Eclairage CIE coordinates of (0.15, 0.05) via the steric effect. Importantly, this represents an advance in deep-blue-emitting fluorescent OLED architectures and materials that meet the requirements of high-definition display.

Introduction

New-generation organic light-emitting diodes (OLEDs) are promising candidates for future displays and solid-state lighting applications owing to their unique cost-effective, lightweight, and flexibility features (Kee et al., 2018, Zhang et al., 2017, Pan et al., 2017, Lian et al., 2018). More importantly, high-performance deep-blue materials, which have a Commission International de L'Eclairage (CIE) coordinate of CIEy< 0.10 and even match the blue standard with the CIE (x, y) coordinates at (0.15, 0.06) of European Broadcasting Union, can provide widened color gamut, which is crucial to full-color displays (Wu et al., 2013, Tang et al., 2015, Tao et al., 2017, Wada et al., 2018). Hence, it is of great significance to develop deep-blue-emitting materials capable of achieving high efficiency and color saturation simultaneously for OLEDs (Liu et al., 2014, Cai et al., 2016, Lee et al., 2017, Shao et al., 2017).

To achieve maximum external quantum efficiency (EQE), the highest priority is to harvest triplet energy to break the internal quantum efficiency (IQE) limit of 25% for singlet spin states (Luo and Aziz, 2010). As the representative molecular frameworks, phosphorescent materials are extensively investigated because 100% IQE can be obtained (Lee et al., 2013). However, the phosphorescent materials containing non-renewable and unevenly distributed heavy metal elements such as Ir, Au, or Pt and the strong quenching effect of the long-lived triplet excitons at high luminescence hinders the commercialization of phosphorescent OLEDs (Tao et al., 2011, Zhang et al., 2016, Kim et al., 2018). On the other hand, although the use of metal-free thermally activated delayed fluorescence (TADF) materials via a reverse intersystem crossing (RISC) has been considered as a promising strategy and important breakthrough for the realization of highly efficient third-generation OLEDs, it still remains difficult to develop highly efficient deep blue materials with a suitably wide band gap (Hirata et al., 2015, Zhang et al., 2014, Hatakeyama et al., 2016). Moreover, the slow RISC process between the lowest triplet excited state (T1) and the lowest singlet excited state (S1) of TADF materials leads to the accumulation of triplet excitons at high luminance. Thus, even with lower electroluminescence (EL) efficiencies, considerable attention is still directed toward fully organic fluorophores such as carbazole (Lin et al., 2009, Li et al., 2015), imidazole (Li et al., 2016, Chen et al., 2017a), pyrene (Wu et al., 2010, Tao et al., 2010), fluorene (Wu et al., 2004, Gao et al., 2010), anthracene, and their derivatives (Wu et al., 2014, Zambianchi et al., 2016, Li et al., 2017a).

Among the variety of deep-blue fluorescent materials, anthracene and its derivatives are considered to be the most promising deep-blue emitters due to their high photoluminescence (PL) quantum yield, tunability of properties by the modification of molecular structure, and high thermal stabilities (Danel et al., 2002, Kim et al., 2008). Using 9,10-di(naphth-2-yl)anthracene (ADN) and its tertiary butyl derivative, 2-t-butyl-9,10-di-(2-naphthyl) anthracene as emitters, Tang et al. reported blue emissions with good chromaticity and luminance efficiency of 3.5 cd A−1 corresponding to the CIE coordinates of (0.15, 0.23) and a half-lifetime of 4,000 hr with an initial light output of 700 cd m−2 (Shi and Tang, 2002). Tang et al. reported an efficient blue-emitting molecule PIAnCN (phenanthroimidazole-anthracene derivative) consisting of cyano-substituted anthracene, and the PIAnCN-based nondoped OLEDs displayed a maximum EQE of 9.44% with CIE coordinates of (0.14, 0.19) (Tang et al., 2018). Recently, new mechanisms by utilizing triplet excitons, such as locally excited (LE) state, charge transfer (CT) state, and triplet-triplet annihilation (TTA), were developed to increase the IQE for fluorescent emitters in blue OLED applications. For example, by employing D-π-A-based phenanthro[9,10-d]imidazole (PI) carbazole hybrid fluorophores 1-(4-(tert-butyl)phenyl)-2-(4-(4-(9-phenyl-9H-carbazol-3-yl)naphthalen-1-yl)phenyl)- 1H-phenanthro[9,10-d]imidazole (TPINCz) with remarkable hybridized local charge transfer (HLCT) excited states as an emissive dopant, Chen et al. obtained a violet blue emission with CIE coordinates of (0.153, 0.059) and a record high EQE of 6.56% ± 0.11% at a brightness of 1,000 cd m−2 (Chen et al., 2017b). In addition, Li et al. synthesized a series of materials 2-(2′-hydroxyphenyl)oxazoles possessing an HLCT excited state, and the doped device reached a high EQE of 7.10% and an excellent color purity with the CIE coordinates of (0.15, 0.08) (Li et al., 2017b). From this viewpoint, the LE state, as an efficient way to use RISC arising from its larger transition moment with a larger orbital overlap, can provide a high radiative transition rate that is related to a high PL efficiency.

In this study, we have synthesized two deep-blue fluorescent emitters trans-9,10-bis(2-butoxyphenyl)anthracene (BBPA) and trans-9,10-bis (2,4-dimethoxyphenyl)anthracene (DMPA) based on 9,10-dibromoanthracene derivatives. These two molecules showed high photoluminescence quantum yields (PLQYs), deep-blue emission, and LE states, which make them suitable for efficient blue OLEDs. The nondoped device bearing BBPA and DMPA exhibited a maximum EQE of 5.7% and 4.46%, respectively. To avoid a lower energy state formation of excimer, we blended the synthesized luminogens into an appropriate host 1,3-bis(N-carbazolyl)benzene (mCP), ADN and 10,10′-bis-(4-fluorophenyl)-3,3′-dimethyl-9,9′-bianthracene (MBAn-(4)-F), respectively. Using BBPA- and DMPA-doped mCP (5.0 wt%) as the emission layer (EML), we have achieved efficient deep-blue OLEDs with maximum EQEs of 6.02% and 8.05% at the corresponding CIE coordinates of (0.15, 0.03) and (0.15, 0.05), respectively. Furthermore, we have found that the use of host matrices with twisted geometric molecular configuration, ADN and MBAn-(4)-F in BBPA emission layer, could significantly improve the intercoupling of blue OLEDs. The device containing the EML of BBPA-doped MBAn-(4)-F exhibited an unprecedentedly high EQE of 10.27% and deep-blue emission with the CIE coordinates of (0.15, 0.05) via the steric effect.

Results and Discussion

Material Synthesis and Characterization

The synthetic routes for the two anthracene-based blue-emitting compounds are shown in Figure 1A. DMPA and BBPA can be synthesized through a one-pot Suzuki cross-coupling reaction from 9,10-dibromoanthracene and the corresponding boronic acid derivatives, with tetrakis(triphenylphosphine)palladium as the catalyst. They were purified by column chromatography on silica gel with dichloromethane and petroleum ether as the eluents. The two products were characterized by 1H and 13C NMR spectroscopy (Figures S1–S4) and high-resolution mass spectrometry. In addition, the structure of BBPA was also confirmed by single-crystal X-ray diffraction analysis (Figure 1C and Table S1).

Figure 1

Molecular Structure Characteristics

(A) Synthetic approach to BBPA and DMPA.

(B) Molecular structures for BBPA and DMPA optimized at the B3LYP/6-1G(d) level of theory using Gaussian 09.

(C) Crystal packing of BBPA molecules.

Also see Figures S1–S4 and Table S1 and Data S1.

To further reveal the electronic structures and excited state properties of these molecules, density functional theory (DFT) and time-dependent DFT calculations at the B3LYP level of theory using the basis set 6-31G(d) were performed. The frontier orbital distributions of this molecule between the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) levels are shown in Figure 1B, and the data are summarized in Table S2. The HOMO and LUMO levels of BBPA and DMPA have similar distributions, being predominantly distributed over the central anthracene core. The energy levels of the HOMO and LUMO were calculated to be in the range -(4.86–4.87) and -(1.37–1.41) eV, respectively, resulting in a wide energy gap of about 3.50 and 3.45 eV.

Photophysical and Thermal Properties

Figure 2A shows the ultraviolet-visible absorption and PL spectra of nondoped solid film of BBPA and DMPA. The spectra in CH2Cl2 solution of the two molecules are shown in Figure S5. The key data are summarized in Table 1. The fluorescence spectra of the nondoped films exhibited maximum emission bands at 433 and 439 nm with a relatively narrow full width at half-maximum (FWHM) of 54 nm and 42 nm, respectively, for BBPA and DPMA. As displayed in Figure 2B, the BBPA and DMPA in nondoped solid films showed monoexponential decay with lifetime (τs) of 3.13 and 2.48 ns, respectively. No delayed component can be observed, indicating that the emission exclusively originates from the prompt decay of the S1 state (Figure S6).

Figure 2

Photophysical Characteristics of BBPA and DMPA

(A) Absorption (black line) and fluorescent (blue line) spectra of BBPA and DMPA.

(B) The corresponding PL decay spectra in solid film at ambient condition.

Also see Figures S5 and S6, and Table S2.

Table 1

Summary of Thermal and Photophysical Properties of BBPA and DMPA

Compound	Tda/Tmb (°C)	λabs, maxc (nm)	λPL, maxd (nm)	PLQYe (%)	HOMO/LUMOf (eV)
BBPA	257/498	399/378/360/341	432	89.5	−5.81/−2.63
DMPA	309/339	402/380/361/341	438	87.0	−5.78/−2.48

Also see Figures S7 and S8.

Decomposition temperature.

Melting point.

Absorption maximum.

Fluorescence emission peak in nondoped thin film.

Absolute PL quantum yield evaluated by integrating sphere of nondoped film.

Determined from the onset of photoelectron spectroscopy in the air, and the equation LUMO = HOMO + Eg.

The thermal stability of the emitters was determined by measurement of their glass transition temperature (Tg), decomposition temperature (Td), and melting point (Tm) using differential scanning calorimetry and thermal gravimetric analysis under a nitrogen atmosphere. Both BBPA and DMPA have shown excellent thermal stability for organic electronics with Td at 257°C and 309°C, respectively (in Figure S7), The photophysical and thermal properties of these two molecules are summarized in Table 1.

Photoelectron yield spectroscopy was employed to measure the practical HOMO levels of the two emitters in the neat thin films (Figure S8). The LUMO level was then determined by adding the corresponding HOMO level and the energy gap (Eg). As listed in Table 1, the HOMO and LUMO levels are estimated to be −5.81 and −2.63 eV for BBPA and −5.78 and −2.48 eV for DMPA, respectively.

Electroluminescence Performance

To evaluate the EL performance of the two materials as emitting cores, we initially fabricated two nondoped OLEDs (devices A0 and B0) with a structure of ITO/HAT-CN (5 nm)/TAPC (40 nm)/EML (20 nm)/Bphen (30 nm)/Liq (1 nm)/Al (150 nm). All the material structures are displayed in Figure S9, and the details of the device fabrication are given in the Methods. The device structures and the electroluminescent properties are shown in Figure 3 and summarized in Table 2.

Figure 3

Performance of Nondoped Deep-Blue OLEDs Based on BBPA and DMPA Emitters.

(A) Device structures.

(B) Current density versus voltage characteristics.

(C) Luminance versus voltage characteristics.

(D–F) (D) External quantum efficiency versus luminance. (E and F) Electroluminescent spectra from BBPA and DMPA fluorescent OLEDs at different driving voltages.

Also see Figure S9.

Table 2

Summary of OLED Characteristics Using BBPA and DMPA as Dopants

Devicea	Emitter	Vonb (V)	ELc (λmax)	CIEd (x, y)	FWHMe (nm)	EQEf (%)
						Max/100/1,000
A0	BBPA	3.2	432	0.15, 0.06	52	5.28/4.02/1.35
B0	DMPA	3.3	436	0.16, 0.08	62	4.97/4.61/4.22
A1	mCP:BBPA	3.5	412	0.15, 0.03	43	6.02/5.14/2.41
B1	mCP:DMPA	3.3	428	0.15, 0.05	51	8.05/8.02/7.16
A2	ADN:BBPA	3.3	432	0.15, 0.06	50	8.67/7.15/8.62
B2	ADN:DMPA	3.7	448	0.15, 0.09	68	4.84/4.58/4.62
A3	MBAn-(4)-F:BBPA	3.6	432	0.15, 0.05	51	10.27/8.70/7.25

Also see Figure S14, and Table S4.

Device configuration: ITO/HAT-CN (5 nm)/TAPC (40 nm)/EML (20 nm)/Bphen (30 nm)/Liq (1 nm)/Al (100 nm).

The operating voltage at a brightness of 1 cd m−2.

The EL emission wavelength at the maximum intensity.

CIE 1931 coordinates at 7 V.

FWHM of EL spectra.

EQE at the maximum value/at 100 cd m−2/at 1,000 cd m−2.

The BBPA- and DMPA-based OLEDs display the highest EQE of 5.28% and 4.97% and deep-blue EL emission with CIE coordinates of (0.15, 0.06) and (0.16, 0.08) (at 7 V), respectively. It is worth noting that the device A0 shows serious efficiency roll-off due to excimeric emission at higher working voltage than device B0, in which a high EQE of 4.22% is still maintained at a practical luminance of 1,000 cd m−2 (Figures 3E, S10, and S11). The devices A0 and B0 show an emission peak at 432 and 436 nm, respectively, which coincides with their nondoped PL spectrum. The EL spectra are comparable to the PL spectra of solid film, indicating that the EL is mainly from the emissive materials (Lukas et al., 2003).

We also fabricated and optimized doped devices (devices A1 and B1) to further improve the efficiency and color purity. The detailed description of device optimization is illustrated in the Supplemental Information. The best result was obtained from the device ITO/HAT-CN (5 nm)/TAPC (40 nm)/mCP: dopants (20 nm)/Bphen (30 nm)/Liq (1 nm)/Al (150 nm) where mCP is the host material (Figures 4A and 4B). The dopant level was optimized to avoid any spectral broadening from aggregation of the excimeric emission and excavate the EL potential of these compounds (3, 5, and 10 wt % of BBPA and DMPA in mCP, Figures S12 and S13 and Table S3). The EL spectra (Figure 4C) and CIE diagram (Figure 4D) of both devices showed deep-blue emission with peaks at 412 and 428 nm, corresponding to the coordinates of (0.15, 0.03) and (0.15, 0.05), respectively. The light-emitting photographs of devices A1 and B1 are displayed in the inset of Figure 4D. All the doped devices show obviously blue-shifted EL spectra compared with the corresponding nondoped devices. It is noteworthy that device A1 with BBPA as the dopant exhibited extraordinary deep blue emission with an FWHM of 43 nm and a high EQE value of 6.02%. The DMPA-based device B1 also showed excellent efficiency of 8.05% with an FWHM of 51 nm. To the best of our knowledge, the EQEs of devices A1 and B1 are the highest among the reported deep-blue devices with CIEy<0.06 (Tang et al., 2018, Hu et al., 2014, Li et al., 2018). Furthermore, device B1 displayed an efficiency roll-up at low luminance and extremely low-efficiency roll-off at high luminance. For example, device B1 showed a maximum EQE of 8.05% (at 165 cd m−2), although the EQE can still be maintained at 7.02% at 1,000 cd m−2, corresponding to only 11.06% decrease compared with the maximum value. There are two possible reasons for the low-efficiency roll-off. First, excellent charge balance of the devices is achieved. It is well known that the charge balance is important for reducing the efficiency roll-off of an OLED device (Figure S14). The second reason is based on the use of doping blue emissive molecules into the mCP matrix, which can effectively remove the excimeric emission in the nondoped emitters (Figure 4C) resulting from the self-quenching of excitons (Lee et al., 2016).

Figure 4

Performance of Optimized Deep-Blue OLEDs

(A) Current density versus voltage characteristics.

(B) External quantum efficiency versus luminance.

(C) Electroluminescent spectra.

(D) Comparison of the CIE coordinates of the EL spectra from two deep-blue fluorescent OLEDs.

Also see Figures S10–S13, and Table S3.

Intriguingly, we found a novel host-guest system that can significantly improve the device performance by utilizing the steric effect. We employed highly steric molecules, ADN and MBAn-(4)-F as the host of BBPA in devices A2 and A3, and achieved an improved efficiency compared with planar-type mCP-based device (A1). As illustrated in Figure 5, devices A2 and A3 show the same emission peak at 432 nm and efficiencies of 8.67% and 10.27%, respectively. It is rational that the improved efficiency can be attributed to the employment of ADN and MBAn-(4)-F with twisted structure having stronger steric hindrance. Moreover, the use of highly steric host can largely reduce the concentration of excitons and thereby suppress intermolecular quenching effect in the solid state, and simultaneously improve the morphological stability that would resist crystallization and morphological transition-induced deterioration during the device operation (Figure S15) (Hung et al., 2017, Young et al., 2002). On the contrary, we found that device B2 using DMPA-doped ADN as an EML exhibited a decreased efficiency compared with device B1 with mCP as the host (Figure S16). The device B3 based on MBAn-(4)-F:5% DMPA emitter was also fabricated, and the key parameters are shown in Table S4. The EL spectra mainly originated from the ADN host emission (444 nm), which can be demonstrated from the consistent PL emission of the ADN molecule. Different emission mechanisms for the two doping systems can be ascribed to the following reasons. (1) The long chains in BBPA molecule facilitate energy transfer from ADN onto the guest molecule BBPA. (2) For the DMPA-doped ADN system, DMPA cannot effectively harvest the excitons without the long chains, leading to the generation of singlet excitons that are mainly formed in ADN.

Figure 5

Performance of Optimized Deep-Blue OLEDs Based on Steric Effect

(A) Current density versus voltage characteristics.

(B) Luminance versus voltage characteristics.

(C) External quantum efficiency versus luminance.

(D) Electroluminescent spectra.

Also see Figures S15–S19.

We note that the efficiency of the device A3 employing MBAn-(4)-F:5% BBPA is significantly higher than that of the traditional fluorescent singlet exciton limit. This is among the best results of deep-blue fluorescent OLEDs with an EQE greater than 10% and CIEy < 0.06. Nevertheless, Figure S17 displays a linear relationship between current density and EL intensity, and the transient EL decay of device A3 is shown in Figure S18, suggesting that TTA does not play a role (Pu et al., 2012). Furthermore, the transient PL spectrum reveals no delayed fluorescence. Nevertheless, the high EQE is in accordance with the LE state conducted by the natural transition orbital analysis and horizontal orientation of the emitting molecules, as shown in Figure S19. Out-coupling efficiency is also an uncertain parameter, and it is possible for the EQE to climb over 25% if the emitting molecules have a horizontal orientation when discussing the origin of high EQEs. The PLQY (ϕ) of the BBPA in neat film was measured to be 89.5%. The classical theoretical estimate for the maximum EQE of fluorescent OLEDs is considered by the equation: EQE = ã×ϕ×çr×çext, where ã represents the charge balance factor (ideally equal to 1), çr represents the probability of formation of an emissive excited state in a recombination event, ϕ is the PLQY of the emissive material, and çext is the out-coupling efficiency (Hung et al., 2017, Wu et al., 2018). Ideally, we assume that the out-coupling efficiency çext should be 25%. Thus, the calculated çr is around 45.8%. This value is far superior to that of the conventional fluorescent materials (çr = 25%). This result supports the fact that molecular orientation plays the key role in reaching high efficiency for the electrofluorescent devices. Finally, it is worth noting that the devices showing low-efficiency roll-off at high luminance are much desired for the practical full color display and lighting applications.

Conclusions

In conclusion, two efficient blue-emitting molecules BBPA and DMPA were designed and synthesized. The doped devices employing mCP as the host show a maximum EQE of 6.02% and 8.05% with a very-low-efficiency roll-off and deep-blue EL with CIE coordinates of (0.15, 0.03) and (0.15, 0.05), respectively. More importantly, we propose a facile and simple strategy of using steric effect to improve the EL performance for BBPA emitter. The introduction of ADN and MBAn-(4)-F with twisted structure having high steric hindrance to prevent intermolecular aggregation and reduce luminescence quenching is beneficial for increasing the device performance without shifting the emission peak. The resulting emitter employing ADN:5% BBPA and MBAn-(4)-F:5% BBPA displays deep-blue EL with CIE coordinates of (0.15, 0.06) and (0.15, 0.05), together with a high EQE of 8.67% and 10.27% and low roll-off at high brightness up to 1,000 cd m−2, respectively. Our results represent state-of-the-art performance for deep-blue OLEDs, and we postulate that this work will open up a new way for highly efficient OLEDs with steric effect to develop deep-blue-emitting molecules.

Limitations of Study

The molecular orientation of the emitting layer is important to improve the efficiency of the organic light-emitting devices. Although the molecular orientation has been tested, it is still difficult to carry out the optical simulation of emitter due to the limitations of the scientific research conditions.

Methods

All methods can be found in the accompanying Transparent Methods supplemental file.