Influence of the D/A ratio of 1,3,5-triphenylbenzene based starburst host materials on blue electrophosphorescent devices: a comparative study

Abstract

A series of starburst host materials, composed of a central 1,3,5-triphenylbenzene (TPB) skeleton with different peripheral modifications were purposefully designed and synthesized by tuning the ratio of p-type carbazole units and n-type diphenylphosphine oxide (DPPO) units from 0 : 3 to 3 : 0. Their optical, electrochemical and thermal properties were investigated systematically. By gradually tuning the D : A ratio of the peripheral groups, the desired ambipolar characteristics were achieved by DCzPOPB and CzDPOPB. As a result, the blue phosphorescent organic light-emitting diode (PhOLED) fabricated with the hosts doped with FIrpic achieved a low turn-on voltage of 3.3 V, a maximum current efficiency of 32.2 cd A⁻¹, a maximum external quantum efficiency of 16.5%, a maximum luminescence of 22 440 cd m⁻², as well as a low efficiency roll-off. Compared to the unipolar analogues, the device performances were significantly improved. In addition, the systematic study of the varied ratios of DPPO and carbazole units revealed the structure–property relationships of these starburst molecules, and a new avenue is proposed for the design of bipolar hosts for blue PhOLEDs.

---

Introduction

Organic light emitting diodes (OLEDs) have gradually been applied in full color displays and solid-state lighting, due to their superior characteristics such as good flexibility, light weight, wide-viewing angles and self-emission.^1–6 Therefore, in order to meet the ever-growing needs, the design and fabrication of materials with excellent emission properties, high efficiencies, low roll-off and long lifetimes is urgent.^7–10 PhOLEDs have thus far played the role of a core technology in achieving high efficiency because they can achieve theoretical internal quantum efficiency (IQE) of 100%, compared to traditional fluorescent OLEDs.^11–13 However, there are still some unfavourable effects, such as concentration quenching and triplet–triplet annihilation (TTA), that usually cause serious efficiency roll-off.^14–20 Thus, developing stable and efficient phosphorescent hosts is urgent and meaningful. Unfortunately, suitable blue phosphorescent hosts are still rare, since they must meet tough conditions. Above all, the triplet energy levels (E_T) of the host should be high enough to prevent back energy transfer.²¹ The blue host materials intrinsically possess a wide band gap between the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) levels, which makes it difficult for carriers being injected from the adjacent layers. Therefore, it is necessary to design hosts with appropriate HOMO and LUMO energy levels for the efficient injection of holes and electrons from adjacent layers.²² Moreover, a balanced hole/electron carrier mobility forming a wide exciton recombination zone within the emitting layer (EML) is a major requirement for high performance devices.²³ Last but not least, the thermal stability of the host materials is helpful to smooth the surface morphology during device fabrication.²⁴ In view of all the factors, it is quite challenging to design blue host materials that meet the requirements of high E_T while conserving appropriate energy levels, as well as good bipolar transport properties and outstanding thermal stability.

Nowadays, plenty of efforts have been made to develop bipolar host materials for high-performance PhOLEDs.^25,26 The general approach involves incorporating donors and acceptors like the carbazole unit or the aromatic phosphine oxide group into a molecule, which can easily result in a lower E_T that would not satisfy the requirements for blue phosphorescence hosts. However, the non-planar molecular structure with a starburst core can not only keep the high E_T of the component groups, but also prevent the close packing of the molecules in the solid state. Carbazole derivatives are most widely employed as blue host materials for PhOLEDs, due to their high triplet energy level, sufficient hole transporting properties and wide energy gap.^27,28 The DPPO group is a strong electron-withdrawing unit that can remarkably increase the electron-injection properties, as reported by our group and others.^29,30 In recent years, starburst molecules have played a key role in most promising functional materials for PhOLEDs and a lot of starburst molecules with novel architectures and sophisticated topologies have been developed.^31–33 For example, a star-shaped bipolar host material based on carbazole and dimesitylboron moieties that possessed high E_T and served as a universal host for efficient red, green and blue electrophosphorescent devices was reported by Shi et al.³⁴ Furthermore, Wong et al. designed a star-shaped molecule named CN-T2T that constructed an exciplex with Tris–PCz. The exciplex ensured balanced charge transport in the emitting layer, achieving a highly efficient green exciplex OLED with external quantum efficiency (EQE) of 11.9%.³⁵ Very recently, three star-shaped compounds consisting of bicarbazolyl side arms and various core moieties were investigated with regard to the influence of the central cores, and all the synthesized compounds exhibited good hole-transporting properties with mobility values as high as 10⁻³ cm² V⁻¹ S⁻¹.³⁶ Nevertheless, it is worth noting that few endeavors have been made to design the bipolar hosts by tuning the D/A ratio of the peripheral groups with the same star-shaped core.

In this contribution, we aim to utilize the scalability of archetypal TPB to elucidate the feasibility of achieving improved blue emitting device performance by tuning the D/A ratio of peripheral groups from 0 : 3 to 3 : 0. 9,9′-(5′-(3-(9H-Carbazol-9-yl)phenyl)-[1,1′:3′,1′′-terphenyl]-3,3′′-diyl)bis(9H-carbazole) (TCzPB), (5′-(3-(9H-carbazol-9-yl)phenyl)-[1,1′:3′,1′′-terphenyl]-3,3′′-diyl)bis(diphenylphosphine oxide) (DCzPOPB), (5′-(3-(9H-carbazol-9-yl)phenyl)-3′′-(9H-carbazol-9-yl)-[1,1′:3′,1′′-terphenyl]-3-yl)diphenylphosphine oxide (CzDPOPB) and (5′-(3-(diphenylphosphoryl)phenyl)-[1,1′:3′,1′′-terphenyl]-3,3′′-diyl)bis(diphenylphosphine oxide) (TPOPB) were conveniently synthesized. Their photophysical properties, such as absorption/emission spectra and low temperature phosphorescence spectra, electrochemical properties and theoretical calculations, were carefully investigated to find the correlation between the structure and luminescence properties. To evaluate the charge transport characters of the four hosts, hole/electron-only devices were fabricated. Finally, blue PhOLEDs were fabricated with the well-known emitter iridium(III)bis(4,6-(di-fluorophenyl)pyridinato-N,C²′)picolinate (FIrpic) (E_T = 2.62 eV),³⁷ because of their high triplet energies (2.78–2.80 eV). Accordingly, tuning the D/A ratio of peripheral groups is an efficient solution to constructing bipolar host materials of high-performance PhOLEDs.

Results and discussion

Synthesis

Scheme 1 depicts the synthetic routes and chemical structures of TCzPB, DCzPOPB, CzDPOPB and TPOPB. The key intermediate 3,3′′-dibromo-5′-(3-bromophenyl)-1,1′:3′,1′′-terphenyl (1) was synthesized according to a previously reported procedure.³⁸ Further selective linking with the carbazole by the traditional Ullmann reaction gave the other important intermediates 9-(3′′-bromo-5′-(3-bromophenyl)-[1,1′:3′,1′′-terphenyl]-3-yl)-9H-carbazole (2), 9,9′-(5′-(3-bromophenyl)-[1,1′:3′,1′′-terphenyl]-3,3′′-diyl)bis(9H-carbazole) (3) and TCzPB. The final products TPOPB, DCzPOPB and CzDPOPB were prepared from 1, 2 and 3 with yields of 45%, 60% and 55%, respectively, by the modified Ni(II) catalyst cross-coupling reaction.³⁹ All the developed host materials were purified by silica gel chromatography and repeated thermal gradient vacuum sublimation, before characterization and device fabrication. The molecular structures were confirmed by ¹H NMR, ¹³C NMR, ³¹P NMR, elemental analysis and mass spectroscopy.

Scheme 1 Synthetic route to the starburst bipolar host materials: TCzPB, DCzPOPB, CzDPOPB and TPOPB.

Thermal properties

The thermal properties of the materials were evaluated by thermal gravimetric analysis (TGA) and differential scanning calorimetry (DSC) at a scanning rate of 10 °C min⁻¹ under a nitrogen atmosphere. The decomposition temperature (T_d, corresponding to 5% weight loss) and glass transition temperature (T_g, by DSC measurements during the second-heating scans) of these starburst host materials are shown in Fig. 1, and the basic parameters are summarized in Table 1. All the designed host materials show excellent thermal stability with the T_d above 450 °C, which could be ascribed to the enhanced rigidity of starburst molecules. Simultaneously, distinct endothermic peaks related to the glass transition appear at 124, 132, 138 and 147 °C for TPOPB, CzDPOPB, DCzPOPB and TCzPB, respectively. No recrystallization or phase transition was observed for all of the compounds upon heating beyond T_g. The improved T_g from TPOPB to TCzPB can be attributed to the increasing number of the bulky carbazole groups, which prevent the easy intermolecular packing and crystallization.⁴⁰ The improved thermal performance makes TPB derivatives superior in the stability of film morphology and suppresses the potential for phase separation during long-time device operation, which is a prerequisite for applications in OLEDs.⁴¹

Fig. 1 The thermal curves of the starburst molecules. Inset: DSC curves of the starburst hosts (heating rate 10 °C min⁻¹).

Table 1 Optical, thermal and electrochemical properties of the starburst host materials

Compounds	λabs,max^a (nm)	λpl,max^a (nm)	Eg^b (eV)	HOMO/LUMO^c (V)		HOMO/LUMO^d (V)		ET^e (eV)	Tg/Td^f (°C)
a Measured in toluene solutions.b Calculated from the onset of the absorption.c Measured from the CV.d Calculated through the density functional theory (DFT).e Calculated according to the phosphorescence spectra.f Measured from the DSC and TGA instruments, respectively.
TPOPB	280	363	3.7	6.1	2.4	6.3	2.4	2.78	124/519
CzDPOPB	340/326/291/280	362/346	3.5	5.7	2.2	5.5	2.2	2.80	132/496
DCzPOPB	340/326/291/280	361/346	3.5	5.8	2.3	5.4	2.3	2.79	138/501
TCzPB	340/326/292/284	361/346	3.5	5.6	2.1	5.4	2.2	2.79	147/459

---

Photophysical properties

Fig. 2(a) displays the absorption and emission spectra of TCzPB, DCzPOPB, CzDPOPB and TPOPB in toluene at room temperature, and the key characteristics are summarized in Table 1. The absorption spectra of these carbazole-based compounds are similar, with the same absorption peaks at 280, 291, 326 and 340 nm, except for TPOPB, which has only one peak around 280 nm. It is reasonable that these absorption peaks are assigned to the π–π* transitions of the TPB core (280 nm) and carbazole (291, 326 and 340 nm). The absorption edge of the UV/Vis spectrum of TPOPB lies at 331 nm, which corresponds to the band gap (E_g) of 3.7 eV, while the other three have a relatively narrow band gap of 3.5 eV. Generally, the band gaps correspond to the degree of conjugation of the molecules, and are approximately equivalent to the singlet excited energy levels (S₁).⁴² A more careful comparison of the optimized structures of the four materials shown in Fig. 4 reveals that the TPB core is non-planar and the peripheral groups, carbazole and DPPO, are meta linked to TPB, which leads to weak charge transfer between the peripheral groups and the core. Therefore, the increase in the electron-donating carbazole unit can only slightly change the degree of conjugation of the molecule, which results in the same E_g for the TCzPB, DCzPOPB and CzDPOPB. However, the phosphorus atom of the DPPO is sp³ hybridized, which blocks the conjugation with the TPB core. Therefore, TPOPB has the smallest conjugation and the largest energy gap. In the PL spectra, TCzPB, DCzPOPB and CzDPOPB have a similar PL emission peak at around 346 nm that could be ascribed to the emission of carbazole, and they all shoulder at around 361 nm, which results from the DPPO unit; the TPOPB exhibits only one peak at 363 nm.

Fig. 2 (a) The absorption and PL spectra of the molecules in toluene solution at 298 K. (b) The phosphorescence spectra at 77 K in 2-MeTHF.

As shown in Fig. 2(b), the phosphorescence spectrum in 2-methyltetrahydrofuran at 77 K shows the E_T (T₁–S₀) of TCzPB, DCzPOPB, CzDPOPB and TPOPB ranging from 2.78 to 2.80 eV. Particularly, the E_T of TPOPB is 2.78 eV, which is the same as the previously reported value.³⁰ As the molecular core of TPB is propeller-shaped, with dihedral angles between 7 and 49°, depending on the substitution and the packing,⁴³ the intramolecular steric interactions are effectively inhibited and the degree of conjugation is not expanded. On the other hand, the DPPO unit improves the charge transport properties without degrading the E_T of the TPB core. Based on the results of the measurements, the newly synthesized compounds with high triplet energy levels could serve as appropriate hosts for FIrpic.

Electrochemical properties

The cyclic voltammograms (CV) were measured in dichloromethane to study the HOMO and LUMO energy levels of the hosts (Fig. 3) and the data are summarized in Table 1. CV analyses indicate that TCzPB, DCzPOPB and CzDPOPB display reversible oxidation waves within the electrochemical window of CH₂Cl₂, which most likely results from the electrochemical properties of the C₃ and C₆ of the carbazole.⁴⁴ Conversely, TPOPB displays high irreversible oxidation potential for the incorporation of DPPO groups, and finally leads to a low-lying HOMO energy level. The HOMO is calculated from the onset of the oxidation potentials by comparing with ferrocene (Fc) (HOMO = −(E_ox + 4.68) eV). In comparison to TPOPB (−6.1 eV), much higher HOMO energy levels of −5.6 to −5.8 eV for TCzPB, DCzPOPB and CzDPOPB were detected, and the LUMO energy levels were in the range of −2.1 to −2.4 eV, which were calculated from the HOMO values and E_g. Three carbazole moieties in TCzPB result in the smallest onset voltage at ∼0.92 V, corresponding to the HOMO of approximately −5.60 eV. The similar oxidation peaks of CzDPOPB and DCzPOPB to TCzPB further demonstrate the restrained intramolecular interplays between the functional groups. As a result of the meta linkage of the carbazole and the insulating linkage of PO moieties,^45–48 the variation in the ratio of carbazole to PO from 1 : 2 to 2 : 1 had little influence on the HOMO energy levels of CzDPOPB and DCzPOPB. TPOPB, which has the most electron-withdrawing DPPO groups, achieved the deepest HOMO energy level of −6.1 eV, which is about 0.5 eV lower than that of TCzPB.

Fig. 3 Cyclic voltammograms of the oxidation processes of the starburst host materials.

Theoretical calculations

To further understand the effect of the topology structure on the electronic properties, density functional theory (DFT) calculations were carried out using B3LYP hybrid functional theory with 6-31G* basis sets. The optimized structures and the HOMO/LUMO distributions for these starburst host materials are given in Fig. 4. The HOMOs for TCzPB, CzDPOPB and DCzPOPB are mainly located on the electron-donating carbazole moiety, while the LUMOs are mainly dispersed in the electron-accepting DPPO unit and the benzenes that link to it directly. For the TPOPB, the HOMO and LUMO are both spread on the TPB core. As we all know, the carbazole is an electron-donating unit, which induces the localization of the HOMO level. The localization of the LUMO in the phenyl units is due to the electron-withdrawing DPPO group, which makes the phenyl unit electron deficient. The separation between the HOMO and LUMO is beneficial to the efficient charge-carrier transport, further reducing the driving voltage of the devices.⁴⁹ Moreover, the trends of the calculated energy levels are conceivably consistent with the experimental ones, although the exact values are somewhat different (see Table 1).

Fig. 4 HOMO and LUMO distributions of the starburst host materials.

Bipolar transport characteristics

Hole-only and electron-only devices of the four host materials TCzPB, DCzPOPB, CzDPOPB, and TPOPB were fabricated to compare hole and electron transporting properties. The hole-only devices (HOD) were designed with the following structure: ITO/NPB (10 nm)/host (30 nm)/NPB (10 nm)/Al (100 nm), whereas the electron-only devices (EOD) were composed of the structure of ITO/TmPyPB (30 nm)/host (30 nm)/TmPyPB (30 nm)/LiF (1 nm)/Al (100 nm). NPB (N,N′-bis-(1-naphthyl)-N,N′-diphenyl-1,10-biphenyl-4,4′-diamine) and TmPyPB (3,3′-(5′-(3-(pyridine-3-yl) phenyl)-[1,1′:3′,1′′-terphenyl]-3,3′′-diyl)dipyridine) layers were used to prevent electron and hole injection from the cathode and anode, respectively. It can be assumed that only hole or electron carriers are injected and transported in the devices, because the LUMO of NPB is high-lying and the HOMO of the TmPyPB is low enough to block electron and hole injection, respectively. As shown in Fig. 5, the current density of the TPOPB based EOD is significantly higher than that of the others, but the current density of the HOD is the lowest. The opposite is true for the TCzPB, which possesses superior hole transport properties compared to electron transport properties. For DCzPOPB and CzDPOPB, the hole and electron current densities are similar over all voltage ranges measured, indicating their bipolar charge transport properties. However, the relatively small difference in the current density between the HOD and EOD of CzDPOPB compared to DCzPOPB, indicated that the charge balance may be improved for the CzDPOPB based emitting device. It seems feasible that ambipolar characteristics could be authentically modulated by tuning the D : A ratio.

Fig. 5 (a) The current density versus voltage curves of the electron-only devices of the four starburst host materials. (b) The current density versus voltage curves of the hole-only devices of the four starburst host materials.

Electrophosphorescent OLED characterization

The bipolar transport characteristics, the excellent morphology stability as well as the high triplet energy suggest that CzDPOPB and DCzPOPB can be used as the ideal hosts for the blue electrophosphorescent devices. FIrpic-doped devices A–D were fabricated with a simple configuration of ITO/MoO₃ (10 nm)/NPB (40 nm)/mCP (5 nm)/host: 6 wt% FIrpic (20 nm)/TmPyPB (40 nm)/LiF (1 nm)/Al (150 nm), (host: TPOPB for device A, CzDPOPB for device B, DCzPOPB for device C and TCzPB for device D), as shown in Fig. 6. In these devices, ITO (indium tin oxide) and Al (aluminum) are the anode and cathode, respectively. NPB is the hole-transporting layer and mCP (1,3-di(9H-carbazol-9-yl)benzene) is the exciton blocking layer. The FIrpic doped host is used as the EML, the best electroluminescence (EL) performance was achieved with 6 wt% FIrpic for all the hosts. TmPyPB was used as both the electron transporting layer (ETL) and hole blocking layer. MoO₃ and LiF (lithium fluoride) served as the hole and electron injecting layers, respectively.

Fig. 6 Schematic of the EL device configurations and the energy levels of the relevant compounds.

Fig. 7(a) shows the current intensity–voltage–luminance characteristics of devices A–D. Evidently, device A, B and C have a similar turn-on voltage (voltage at 1.0 cd m⁻²) of about 3.3 V, which is lower than for the device D. The high turn-on voltage of device D could be ascribed to the large electron injection barrier from TmPyPB to TCzPB. It is known that holes are the major carriers and that the Ir³⁺ complexes (FIrpic) can assist hole injection and transportation through charge capture.^50–52 For device A, B and C, Ir³⁺ complexes may play an important role in their low turn-on voltage because of the existence of the electron transporting group, DPPO. The relatively high current density of devices A and D indicate that the compounds TPOPB and TCzPB possess good electron-transporting and hole-transporting properties, respectively. However, the luminance of the exciton recombination of devices B and C is apparently higher than devices A and D, which can be attributed to the improved charge balance by the bipolar nature of CzDPOPB and DCzPOPB.

Fig. 7 (a) The current density–voltage–luminescence characteristics of devices A–D. (b) The power efficiency–current density–EQE of device A–D. (c) The EL spectra of the devices A–D at a voltage of 8 V.

Fig. 7(b) exhibits the power efficiency–current intensity–EQE of device A–D. The maximum luminous efficiency increases from 14.6 lm W⁻¹ for TPOPB, 16.2 lm W⁻¹ for TCzPB to 25.7 lm W⁻¹ for DCzPOPB, 29.1 lm W⁻¹ for CzDPOPB. Correspondingly, the external quantum efficiencies (EQE) of 16.5% for CzDPOPB and 16.4% for DCzPOPB are higher than that of TCzPB (11.7%), and nearly two times that of TPOPB (9.0%). Furthermore, it is found that even at a brightness of 1000 cd m⁻², device B and device C still show the EQE as high as 15.2% and 15.0%, respectively. The obtained small roll-off is presumably due to the conductivity of holes and electrons simultaneously through the whole EML. For the bipolar hosts, the excitons are believed to be formed in the broader EML, rather than located at the narrow interface of EML/ETL, so that the density of the triplet excitons can be controlled at a relatively low level to minimize the T₁–T₁ annihilation, and thus prohibit the steep efficiency roll-off. In addition, the strong steric effect of these host materials is beneficial to suppressing TTA for high efficiencies.⁵³ The bipolar transporting properties of the CzDPOPB and DCzPOPB and the well matched energy levels with the neighboring functional layers are critical factors for the high efficiencies and low roll-off of blue PhOLEDs.

The EL spectra of the devices are shown in Fig. 7(c). All the EL spectra of the devices A–D exhibit a similar CIE coordinate around (0.15, 0.30), corresponding to the emission of FIrpic. Moreover, no additional emission from the starburst hosts is observed, which demonstrates that the exciton energy can be entirely transferred from the host materials to FIrpic (Table 2).

Table 2 EL data of the devices A–D

Device	Host	Von^a (V)	Lmax [cd m^−2] (V at Lmax, V)	ηc^b [cd A^−1]	ηp^c [lm W^−1]	EQE^d [%]	CIE [x, y]^e
a Von: the voltage of 1 cd m^−2, V: the voltage of the maximum brightness.b Maximum ηc and ηc at 1000 cd m^−2.c Maximum ηp and ηp at 1000 cd m^−2.d Maximum external quantum efficiency.e Measured at 8.0 V.
A	TPOPB	3.4	3082 (9.7)	16.7/5.2	14.6/3.7	9.0	(0.15, 0.31)
B	CzDPOPB	3.3	14 060 (11.5)	32.2/23.7	29.1/14.7	16.5	(0.15, 0.31)
C	DCzPOPB	3.3	22 440 (12.0)	32.1/20.6	25.7/14.5	16.4	(0.15, 0.30)
D	TCzPB	3.6	6537 (10.8)	21.8/18.5	16.2/8.8	11.7	(0.15, 0.30)

---

Conclusion

In summary, a series of new starburst host materials with gradual modulation of the ratio between carbazole and DPPO units from 0 : 3 to 3 : 0 were designed and synthesized. Their excellent thermal properties, high E_T (around 2.80 eV) and bipolar transporting characteristics were carefully tuned, and stable performances of blue electrophosphorescent FIrpic-doped devices with comparable efficiency (16.5%) have been realized using CzDPOPB as host. It should be noted that there is still room for enhancing the device performance for practical applications. Our study offers a promising avenue to obtain bipolar hosts with high triplet energies based on the star-shaped scaffold, for application in efficient and stable blue PhOLEDs.

Experimental section

Materials and measurements

All the reagents and solvents used for the synthesis were purchased from Aldrich and used without further purification. All reactions were performed under a dry nitrogen atmosphere. ¹H NMR and ¹³C NMR spectra were acquired on a Bruker-AF301 AT 400 MHz spectrometer. Elemental analyses of carbon, hydrogen, and nitrogen were performed on an Elementar (Vario Micro cube) analyzer. Mass spectra were carried out on an Agilent (1100 LC/MSD Trap) using ACPI ionization. UV-Vis absorption spectra were recorded on a Shimadzu UV-VIS-NIR Spectrophotometer (UV-3600). PL spectra were recorded on Edinburgh instruments (FLSP920 spectrometers). Differential scanning calorimetry (DSC) was performed on a PE Instrument, DSC 2920 unit at a heating rate of 10 °C min⁻¹ from 30 to 300 °C under nitrogen. The glass transition temperature (T_g) was determined from the second heating scan. Thermogravimetric analysis (TGA) was undertaken with a PerkinElmer Instrument (Pyris1 TGA). The thermal stability of the samples under a nitrogen atmosphere was determined by measuring their weight loss while heating at a rate of 10 °C min⁻¹ from 30 to 700 °C. Cyclic voltammetry measurements were carried out in a conventional three-electrode cell using a Pt carbon working electrode of 2 mm in diameter, a platinum wire counter electrode, and an Ag/AgNO₃ (0.1 M) reference electrode on a computer-controlled EG&G Potentiostat/Galvanostat model 283 at room temperature. CV of all compounds was performed in dichloromethane containing 0.1 M tetrabutylammoniumhexafluorophosphate (Bu₄NPF₆) as the supporting electrolyte. The onset potential was determined from the intersection of two tangents drawn at the rising and background current of the cyclic voltammogram. All solutions were purged with a nitrogen stream for 10 min before measurement. The procedure was performed at room temperature and a nitrogen atmosphere was maintained over the solution during measurements. The experimental conditions and equipment used have been described in our previous works.⁵⁴

Computational details

The geometrical and electronic properties were determined with the Amsterdam Density Functional (ADF) 2009.01 program package. The calculation was optimized by means of the B3LYP (Becke three parameters hybrid functional with Lee–Yang–Perdew correlation functional),⁵⁵ with the 6-31G (d) atomic basis set. The electronic structures were calculated at τ-HCTHhyb/6-311++G (d, p) level.⁵⁶ Molecular orbitals were visualized using ADF view.

Device fabrication and measurement

MoO₃, NPB, mCP, and TmPyPB are commercially available. Commercial ITO coated glass with sheet resistance of 20 Ω per square was used as the starting substrate. Before device fabrication, the ITO glass substrates were pre-cleaned carefully and treated with oxygen plasma for 5 min. The sample was then transferred to the deposition system. MoO₃ (10 nm) was deposited first on the ITO substrate, followed by NPB (40 nm), mCP (5 nm), EML (20 nm), and TmPyPB (40 nm). Finally, a cathode composed of lithium fluoride (1 nm) and aluminum (150 nm) was sequentially deposited onto the substrate under vacuum of 10⁻⁶ Torr. The J–V–L of the devices was measured with a Keithley 2400 Source meter equipped with a calibrated silicon photodiode. The EL spectra were measured by a PR655 spectrometer. The EQE values were calculated according to previously reported methods.⁵⁴ All measurements were carried out at room temperature under ambient conditions.

Synthesis

Compounds 2, 3 and TCzPB were synthesized through the following procedure: a mixture of carbazole, 3,3′′-dibromo-5′-(3-bromophenyl)-1,1′:3′,1′′-terphenyl (1) with different ratios (1 : 1, 1 : 2 and 1 : 3), CuI, 18-crown-6, and K₂CO₃ in 1,3-dimethyl-3,4,5,6-tetrahydro-2(1H)-pyrimidinone (DMPU) was heated at 170 °C for 12 h under nitrogen. After cooling to room temperature, dichloromethane was added and the mixture was filtered. The solvent was removed under reduced pressure and the residue was purified by column chromatography on silica gel, with dichloromethane as eluent to give a white powder. TPOPB, CzDPOPB and DCzPOPB were synthesized according to the following procedure: to a flask containing NiCl₂·6H₂O, zinc, 2,2′-bipyridine, diphenylphosphine oxide and dimethylacetamide (DMAc), 2 or 3 were added. The reaction mixture was then stirred with a stir bar at 110 °C for 24 h. After completion of the reaction, the mixture was allowed to cool to room temperature and water added. The organic layer was isolated and the remaining aqueous phase was further extracted with CH₂Cl₂. The organic phases were then combined and dried with anhydrous MgSO₄, and purified by silica gel column chromatography using dichloromethane–methanol as the eluent to afford the corresponding product.

9-(3′′-Bromo-5′-(3-bromophenyl)-[1,1′:3′,1′′-terphenyl]-3-yl)-9H-carbazole (2). Yield: 40%. ¹H NMR (CDCl₃, 400 MHz): δ (ppm) 8.17–8.15 (d, J = 7.6 Hz, 2H), 8.09–8.07 (m, 4H), 7.89–7.79 (m, 4H), 7.78–7.77 (t, J = 1.4 Hz, 1H), 7.73–7.71 (d, J = 7.6 Hz, 1H), 7.69–7.40 (m, 10H), 7.39–7.22 (m, 9H). ¹³C NMR (CDCl₃, 100 MHz): δ (ppm) 142.75, 142.63, 141.74, 141.36, 140.90, 139.46, 138.41, 130.74, 130.48, 130.43, 130.35, 126.49, 126.43, 126.03, 125.99, 125.83, 125.73, 125.58, 123.42, 123.34, 123.05, 120.38, 120.33, 120.03, 119.43, 110.56, 109.75. MS (APCI): calcd for C₄₈H₃₁N₂Br: 715.7, found, 716.4 (M + 1)⁺.

9,9′-(5′-(3-Bromophenyl)-[1,1′:3′,1′′-terphenyl]-3,3′′-diyl)bis(9H-carbazole) (3). Yield: 35%. ¹H NMR (CDCl₃, 400 MHz): δ (ppm) 8.18–8.16 (d, J = 7.6 Hz, 2H), 7.89 (s, 1H), 7.83–7.79 (m, 5H), 7.73–7.72 (m, 2H), 7.63–7.59 (m, 3H), 7.53–7.47 (m, 4H), 7.45–7.41 (m, 2H), 7.36–7.29 (m, 4H). ¹³C-NMR (CDCl₃, 100 MHz): δ (ppm) 142.73, 142.58, 141.64, 141.28, 140.90, 138.45, 130.74, 130.46, 130.41, 130.33, 126.47, 126.38, 126.03, 125.95, 125.59, 125.48, 123.46, 123.06, 120.38, 120.06, 109.74. MS (APCI): calcd for C₃₆H₂₃NBr₂: 629.4, found, 630.1 (M + 1)⁺.

9,9′-(5′-(3-(9H-Carbazol-9-yl)phenyl)-[1,1′:3′,1′′-terphenyl]-3,3′′-diyl)bis(9H-carbazole) (TCzPB). Yield: 80%. ¹H-NMR (CDCl₃, 400 MHz): δ (ppm) 8.15–8.13 (d, J = 7.6 Hz, 6H), 7.89 (s, 6H), 7.80–7.78 (d, J = 7.6 Hz, 3H), 7.71–7.68 (m, 3H), 7.60–7.58 (d, J = 8.0 Hz, 3H), 7.46–7.43 (d, J = 8.4 Hz, 6H), 7.40–7.36 (m, 6H), 7.30–7.25 (m, 6H). ¹³C-NMR (CDCl₃, 100 MHz): δ (ppm) 142.67, 141.83, 140.93, 138.43, 130.47, 126.50, 126.45, 126.08, 126.01, 125.70, 123.43, 120.35, 120.01, 109.74. MS (APCI): calcd for C₆₀H₃₉N₃: 801.3, found, 802.3 (M + 1)⁺. Elemental analysis calcd (%) for C₆₀H₃₉N₃: C, 89.86; H, 4.90; N, 5.24; found: C, 89.70; H, 4.99; N, 5.10.

(5′-(3-(9H-Carbazol-9-yl)phenyl)-3′′-(9H-carbazol-9-yl)-[1,1′:3′,1′′-terphenyl]-3-yl)diphenylphosphine oxide (CzDPOPB). Yield: 55%. ¹H NMR (CDCl₃, 400 MHz): δ (ppm) 8.18–8.17 (d, J = 7.6 Hz, 2H), 8.06–8.02 (m, 2H), 7.84–7.82 (m, 2H), 7.75–7.66 (m, 14H), 7.60–7.51 (m, 5H), 7.48–7.39 (m, 16H), 7.32–7.28 (m, 2H). ¹³C NMR (CDCl₃, 100 MHz): δ (ppm) 142.46, 141.10, 140.98, 138.36, 132.89, 132.17, 132.07, 131.85, 131.35, 131.25, 131.02, 130.92, 130.84, 130.47, 129.08, 128.96, 128.50, 126.51, 126.06, 125.65, 123.42, 120.38, 120.05, 109.80. ³¹P NMR (CDCl₃, 400 MHz): δ (ppm) 29.26. MS (APCI): calcd for C₆₀H₄₃NO₂P₂: 871.3, found, 872.3 (M + 1)⁺. Elemental analysis calcd (%) for C₆₀H₄₃NO₂P₂: C 82.65, H 4.97, N 1.61; found: C 82.64, H 4.98, N 1.64.

(5′-(3-(9H-Carbazol-9-yl)phenyl)-[1,1′:3′,1′′-terphenyl]-3,3′′-diyl)bis(diphenylphosphine oxide) (DCzPOPB). Yield: 60%. ¹H NMR (CDCl₃, 400 MHz): δ (ppm) 8.17–8.15 (d, J = 7.6 Hz, 4H), 8.01–7.98 (d, J = 12.4 Hz, 1H), 7.87–7.82 (m, 4H), 7.78–7.76 (m, 4H), 7.72–7.63 (m, 6H), 7.62–7.54 (m, 4H), 7.45–7.26 (m, 18H). ¹³C NMR (CDCl₃, 100 MHz): δ (ppm) 142.57, 141.61, 141.02, 140.90, 138.40, 132.87, 132.14, 132.01, 131.83, 131.35, 131.08, 130.98, 130.48, 129.15, 129.02, 128.57, 128.45, 126.54, 126.04, 125.65, 123.43, 120.37, 120.03, 109.77. ³¹P NMR (CDCl₃, 400 MHz): δ (ppm) 29.27. MS (APCI): calcd for C₆₀H₄₁N₂OP: 836.3, found, 837.7 (M + 1)⁺. Elemental analysis calcd (%) for C₆₀H₄₁N₂OP: C 86.10, H 4.94, N 3.35; found: C 86.12, H 4.93, N 3.34.

(5′-(3-(Diphenylphosphoryl)phenyl)-[1,1′:3′,1′′-terphenyl]-3,3′′-diyl)bis(diphenyl phosphine oxide) (TPOPB). Yield: 45%. ¹H NMR (CDCl₃, 400 MHz): δ (ppm) 8.08–8.05 (d, J = 12.4 Hz, 3H), 7.81–7.79 (m, 3H), 7.73–7.68 (m, 15H), 7.56–7.46 (m, 24H). ¹³C NMR (CDCl₃, 100 MHz): δ (ppm) 141.53, 141.17, 141.05, 133.89, 132.97, 132.86, 131.94, 131.22, 130.81, 129.00, 128.87, 125.78. ³¹P NMR (CDCl₃, 400 MHz): δ (ppm) 29.25. MS (APCI): calcd for C₆₀H₄₅O₃P₃: 906.3, found, 907.3 (M + 1)⁺. Elemental analysis calcd (%) for C₆₀H₄₅O₃P₃: C, 79.46; H, 5.00; found: C 79.58, H 5.08.

Acknowledgements

This research work was supported by the science and technology support program of Hubei Province (2015BAA075), the NSFC/China (51573065), the National Basic Research Program of China (973 Program 2013CB922104) and the Analytical and Testing Centre at Huazhong University of Science and Technology.

References

1. L. Xiao, Z. Chen, B. Qu, J. Luo, S. Kong, Q. Gong and J. Kido, Adv. Mater., 2011, 23, 926–952.
2. T. Higuchi, H. Nakanotani and C. Adachi, Adv. Mater., 2015, 27, 2019–2023.
3. M. Zhu and C. Yang, Chem. Soc. Rev., 2013, 42, 4963–4976.
4. L. Ying, C. L. Ho, H. Wu, Y. Cao and W. Y. Wong, Adv. Mater., 2014, 26, 2459–2473.
5. X. Yang, G. Zhou and W. Y. Wong, J. Mater. Chem. C, 2014, 2, 1760–1778.
6. S. R. Forrest, Nature, 2004, 428, 911–918.
7. K. S. Yook and J. Y. Lee, Adv. Mater., 2012, 24, 3169–3190.
8. H. Liu, G. Cheng, D. Hu, F. Shen, Y. Lv, G. Sun, B. Yang, P. Lu and Y. Ma, Adv. Funct. Mater., 2012, 22, 2830–2836.
9. C. Murawski, K. Leo and M. C. Gather, Adv. Mater., 2013, 25, 6801–6827.
10. Z. An, C. Zheng, Y. Tao, R. Chen, H. Shi, T. Chen, Z. Wang, H. Li, R. Deng, X. Liu and W. Huang, Nat. Mater., 2015, 14, 685–690.
11. C. Adachi, M. A. Baldo, M. E. Thompson and S. R. Forrest, J. Appl. Phys., 2001, 90, 5048–5051.
12. X. Yang, G. Zhou and W. Y. Wong, Chem. Soc. Rev., 2015, 44, 8484–8575.
13. H. Xu, R. Chen, Q. Sun, W. Lai, Q. Su, W. Huang and X. Liu, Chem. Soc. Rev., 2014, 43, 3259–3302.
14. E. Mondal, W. Y. Hung, Y. H. Chen, M. H. Cheng and K. T. Wong, Chem.–Eur. J., 2013, 19, 10563–10572.
15. D. Yu, F. Zhao, C. Han, H. Xu, J. Li, Z. Zhang, Z. Deng, D. Ma and P. Yan, Adv. Mater., 2012, 24, 509–514.
16. S. Gong, Y. Chen, J. Luo, C. Yang, C. Zhong, J. Qin and D. Ma, Adv. Funct. Mater., 2011, 21, 1168–1178.
17. C. W. Lee and J. Y. Lee, Adv Mater., 2013, 25, 596–600.
18. C. J. Zheng, J. Ye, M. F. Lo, M. K. Fung, X. M. Ou, X. H. Zhang and C. S. Lee, Chem. Mater., 2012, 24, 643–650.
19. B. Pan, B. Wang, Y. Wang, P. Xu, L. Wang, J. Chen and D. Ma, J. Mater. Chem. C, 2014, 2, 2466–2469.
20. A. Chaskar, H. F. Chen and K. T. Wong, Adv. Mater., 2011, 23, 3876–3895.
21. C. L. Ho and W. Y. Wong, New J. Chem., 2013, 37, 1665–1683.
22. Y. X. Zhang, L. Zhang, L. S. Cui, C. H. Gao, H. Chen, Q. Li, Z. Q. Jiang and L. S. Liao, Org. Lett., 2014, 16, 3748–3751.
23. W. Jiang, Z. Ge, P. Cai, B. Huang, Y. Dai, Y. Sun, J. Qiao, L. Wang, L. Duan and Y. Qiu, J. Mater. Chem., 2012, 22, 12016–12022.
24. P. Kumar, K. N. Shivananda, W. Zajączkowski, W. Pisula, Y. Eichen and N. Tessler, Adv. Funct. Mater., 2014, 24, 2530–2536.
25. Y. H. Son, Y. J. Kim, M. J. Park, H.-Y. Oh, J. S. Park, J. H. Yang, M. C. Suh and J. H. Kwon, J. Mater. Chem. C, 2013, 1, 5008–5014.
26. L. Duan, J. Qiao, Y. Sun and Y. Qiu, Adv. Mater., 2011, 23, 1137–1144.
27. W. C. Lin, H. W. Lin, E. Mondal and K. T. Wong, Org. Electron., 2015, 17, 1–8.
28. Y. Zhao, C. Wu, P. Qiu, X. Li, Q. Wang, J. Chen and D. Ma, ACS Appl. Mater. Interfaces, 2016, 8, 2635–2643.
29. B. Wang, X. Lv, B. Pan, J. Tan, J. Jin and L. Wang, J. Mater. Chem. C, 2015, 3, 11192–11201.
30. S. Gong, Y. L. Chang, K. Wu, R. White, Z. H. Lu, D. Song and C. Yang, Chem. Mater., 2014, 26, 1463–1470.
31. J. Jia, L. Zhu, Y. Wei, Z. Wu, H. Xu, D. Ding, R. Chen, D. Ma and W. Huang, J. Mater. Chem. C, 2015, 3, 4890–4902.
32. W. Jiang, X. Ban and Y. Su, RSC Adv., 2015, 5, 66994–67000.
33. T. Matulaitis, N. Kostiv, J. V. Grazulevicius, L. Peciulyte, J. Simokaitiene, V. Jankauskas, B. Luszczynska and J. Ulanski, Dyes Pigm., 2016, 127, 45–58.
34. H. Shi, D. Xin, X. Dong, J. X. Dai, X. Wu, Y. Miao, L. Fang, H. Wang and M. M. F. Choi, J. Mater. Chem. C, 2014, 2, 2160–2168.
35. W. Y. Hung, P. Y. Chiang, S. W. Lin, W. C. Tang, Y. T. Chen, S. H. Liu, P. T. Chou, Y. T. Hung and K. T. Wong, ACS Appl. Mater. Interfaces, 2016, 8, 4811–4818.
36. N. A. Kukhta, D. Volyniuk, J. V. Grazulevicius and G. Juska, J. Phys. Chem. C, 2016, 120, 1208–1217.
37. Y. Kawamura, K. Goushi, J. Brooks, J. J. Brown, H. Sasabe and C. Adachi, Appl. Phys. Lett., 2005, 86, 071104.
38. V. S. Iyer, M. Wehmeier, J. D. Brand, M. A. Keegstra and K. Müllen, Angew. Chem., Int. Ed., 1997, 36, 1604–1607.
39. H. L. X. Zhang, X. Hu, G. Tang, J. Zhu and Y. Zhao, Org. Lett., 2011, 13, 3478–3481.
40. X. Wang, S. Wang, Z. Ma, J. Ding, L. Wang, X. Jing and F. Wang, Adv. Funct. Mater., 2014, 24, 3413–3421.
41. H. F. Chen, T. C. Wang, W. Y. Hung, H. C. Chiu, C. Yun and K. T. Wong, J. Mater. Chem., 2012, 22, 9658–9664.
42. C. Han, Z. Zhang, H. Xu, J. Li, G. Xie, R. Chen, Y. Zhao and W. Huang, Angew. Chem., Int. Ed., 2012, 51, 10104–10108.
43. H. Detert, M. Lehmann and H. Meier, Materials, 2010, 3, 3218–3330.
44. J. He, H. Liu, Y. Dai, X. Ou, J. Wang, S. Tao, X. Zhang, P. Wang and D. Ma, J. Phys. Chem. C, 2009, 113, 6761–6767.
45. Z. Zhang, Z. Zhang, R. Chen, J. Jia, C. Han, C. Zheng, H. Xu, D. Yu, Y. Zhao, P. Yan, S. Liu and W. Huang, Chem.–Eur. J., 2013, 19, 9549–9561.
46. C. Han, F. Zhao, Z. Zhang, L. Zhu, H. Xu, J. Li, D. Ma and P. Yan, Chem. Mater., 2013, 25, 4966–4976.
47. C. Han, L. Zhu, F. Zhao, Z. Zhang, J. Wang, Z. Deng, H. Xu, J. Li, D. Ma and P. Yan, Chem. Commun., 2014, 50, 2670–2672.
48. D. Ding, Z. Zhang, Y. Wei, P. Yan and H. Xu, J. Mater. Chem. C, 2015, 3, 11385–11396.
49. G. Mu, W. Zhang, P. Xu, H. Wang, Y. Wang, L. Wang, S. Zhuang and X. Zhu, J. Phys. Chem. C, 2014, 118, 8610–8616.
50. C. Han, Z. Zhang, H. Xu, G. Xie, J. Li, Y. Zhao, Z. Deng, S. Liu and P. Yan, Chem.–Eur. J., 2013, 19, 141–154.
51. J. Zhang, D. Ding, Y. Wei, F. Han, H. Xu and W. Huang, Adv. Mater., 2016, 28, 479–485.
52. C. Han, L. Zhu, J. Li, F. Zhao, Z. Zhang, H. Xu, Z. Deng, D. Ma and P. Yan, Adv. Mater., 2014, 26, 7070–7077.
53. J. Zhang, D. Ding, Y. Wei and H. Xu, Chem. Sci., 2016, 7, 2870–2882.
54. B. Wang, G. Mu, J. Tan, Z. Lei, J. Jin and L. Wang, J. Mater. Chem. C, 2015, 3, 7709–7719.
55. C. Lee, W. Yang and R. G. Parr, Phys. Rev. B: Condens. Matter Mater. Phys., 1988, 37, 785–789.
56. A. D. Boese and N. C. Handy, J. Chem. Phys., 2002, 116, 9559–9569.

---