Recent progress of sulphur-containing high-efficiency organic light-emitting diodes (OLEDs)

Abstract

In the last few decades, organic light-emitting diodes (OLEDs) have been rapidly developed and occupy an important position in the lighting and display market. Early conventional fluorescent OLED materials can only utilize 25% of excitons. Improving the harvest of non-emissive triplets has become the most important issue in this field. Organic electroluminescent materials have been developed from the first generation of fluorescent materials to the second generation of phosphorescent materials, and to the new generation of luminogens, mainly including triplet–triplet annihilation (TTA) materials, thermally activated delayed fluorescence (TADF) materials and hot exciton materials. The new generation of luminogens is mainly constructed using heterocyclic compounds with a D–A structure. Sulphur-fused heterocyclic derivatives are some of the most investigated heterocyclic compounds. Sulphur possesses two lone pairs of electrons and an empty d orbital with an outermost electron structure of 3s²3p⁴. Sulfoxide or sulfone with electron-deficient characteristics can be easily formed by oxidation of sulphur, which is beneficial to electron injection and transportation. All these features make sulphur-fused heterocyclic derivatives an appealing building block for the new generation of luminogens. Herein, recent progress of efficient OLEDs based on sulphur-containing heterocyclic compounds, involving thiophene, phenylsulfone, thiazole and phenothiazine, as the active layers, is systematically summarized. The fundamental principles regarding molecular design are discussed, and representative reported TADF and hot exciton sulphur-based luminogens are summarized and analyzed, along with their structure–property relationships and OLED applications. As a structure/property/device performance relationship map is drawn, it is hoped that the review indicates one way to obtain better performance of this series of promising OLEDs.

---

Zijun Feng

Zijun Feng received his bachelor's degree in Chemistry from Jilin University in 2019. He is now a PhD candidate at the State Key Laboratory of Supramolecular Structure and Materials, College of Chemistry, Jilin University. His research interest is focused on light-emitting materials of organic light-emitting diodes.

Zhuang Cheng

Zhuang Cheng received his bachelor's degree in Chemistry from Jilin University in 2020. He is now a master's degree candidate at the State Key Laboratory of Supramolecular Structure and Materials, College of Chemistry, Jilin University. His current scientific interest is focused on the device physics of organic light-emitting materials and devices.

Haixu Jin

Haixu Jin received his bachelor's degree in Chemistry from Zhengzhou University in 2019. He is now a master's degree candidate at the State Key Laboratory of Supramolecular Structure and Materials, College of Chemistry, Jilin University. His current scientific interest is ed on blue organic light-emitting materials and devices.

Ping Lu

Ping Lu received her PhD degree in polymer chemistry and physics from Jilin University in 2005. And then, she joined the State Key Lab of Supramolecular Structure and Materials, Jilin University. She worked as a postdoctoral research fellow at the Hong Kong University of Science and Technology from 2009 to 2010. She was promoted to be a full professor in 2014 at Jilin University. Her current research interests are focused on organic/polymer optoelectronic materials and devices.

---

1. Introduction

Since the pioneering work of organic light-emitting diodes (OLEDs) based on vacuum evaporation reported by Tang and Van Slyke in 1987, OLEDs have exhibited promising prospects and aroused extensive attention.¹ In the past few decades, OLEDs have been gradually developed and occupy an important position in the lighting and display market, owing to their advantages of self-illumination, high brightness, fast response, flexibility being easy to process, etc.^2,3 According to the spin statistical theory, electrically excited electrons would produce 25% singlet excitons and 75% triplet excitons, and only the singlet excitons can be utilized for fluorescence, which results in a limited internal quantum efficiency (IQE) of up to 25% for the first-generation of luminescent materials. Moreover, with an out-coupling efficiency of 20%, the maximum external quantum efficiency (EQE) is only 5%, which is too low to put into practical application.⁴ To utilize non-emissive triplet excitons, heavy-metal atoms were introduced and coordinated with organic molecules, which facilitates phosphorescence emission in OLEDs.^5–10 Since the ISC process can be significantly improved by heavy-metal atoms by enhancing the spin orbital coupling (SOC) effect, the second-generation of phosphorescent OLED (PHOLED) materials could achieve a theoretical maximum IQE of 100%.¹¹ Although the PHOLEDs with an excellent maximum EQE spring up like mushrooms, there are still core issues that need to be addressed: the extremely serious triplet–triplet annihilation (TTA) effect under high operating current density, high preparation cost, and deficient device lifetime.¹² In recent years, a great deal of endeavors have been devoted to the OLEDs based on pure organic small molecules which can utilize triplet excitons. The main mechanisms for harvesting triplet energy include TTA,^13–15 hot excitons,^16–18 and thermally activated delayed fluorescence (TADF).^19,20 TTA is a characteristic mechanism to up-convert the triplet excited states with low energy into singlet excited states (Fig. 1). To accomplish this up-conversion process, two excited T excitons generate one excited S exciton and one exciton in the ground state by a fusion and RISC process. Jortner et al.²¹ proposed that when two triplet excitons are annihilated to generate an intermediate state ((TT)**), there is a 1/9 probability that they would form a singlet intermediate state (¹(TT)**), 3/9 probability that they would form a triplet intermediate state (³(TT)**), and 5/9 probability that they would generate a quintet intermediate state (⁵(TT)**). ¹(TT)** can be converted to S₁ and S₀, ³(TT)** can be converted to T₁ and S₀, and ⁵(TT)** can be converted to T₁ and S₀ or two T₁. Theoretically, when all the intermediate states are transformed into singlet excitons during the TTA process, the total IQE can reach 62.5% with an up-conversion efficiency of 50%. Furthermore, assume that the devices’ optical out-coupling is 20%, the maximum EQE of devices based on the TTA mechanism could reach 12.5%, and this is a significant improvement over the 5% EQE of traditional fluorescent OLEDs.²²

Fig. 1 Photophysical mechanism of TTA.

However, an EQE of 12.5% is not enough to satisfy the demand for higher device performance. Hence, tremendous efforts have been devoted to TADF emitters, which theoretically could achieve a maximum IQE of 100% (Fig. 2). The key luminescence mechanism of TADF is as follows: under the premise that the exchange energy (ΔE_ST) between the lowest triplet excited state (T₁) and the lowest singlet excited state (S₁) is small enough, the T₁ state flips to the S₁ state through the reverse intersystem crossing (RISC) phenomenon at room temperature, and then the S₁ state emits delayed fluorescence through the radiation transition.²³ Therefore, a TADF emitter can be achieved by designing molecules with twisted geometries in which the spatial overlap between the highest occupied molecular orbital (HOMO) level and lowest unoccupied molecular orbital (LUMO) level is small enough to result in a small ΔE_ST because the interchange interaction integral of ΔE_ST and the spatial wave function between the frontier molecular orbitals (FMOs) is inversely proportional.²⁴

Fig. 2 Photophysical mechanism of typical TADF.

However, TADF-based OLEDs encounter serious roll-offs or unsatisfactory operation lifetimes of devices caused by the long triplet lifetime. Different from TADF, triplet excitons can be harvested by the conversion process from triplet states T_n (n ≥ 2) to singlet states S_m (m ≥ 1), which is named the “hot exciton” process (Fig. 3). The hot exciton mechanism allows RISC to occur between the singlet and triplet excited states at higher energy levels, which enables a high efficiency and short triplet excited state lifetime to be achieved simultaneously.²⁵ In early studies, it was found that RISC from higher triplet states (hRISC) exists in organic molecules such as naphthalene, quinoline, isoquinoline, fluorene, rose bengal, tetraphenylporphyrin, cyanine dyes, anthracene and its derivatives, and some excited-state intramolecular proton transfer (ESIPT) materials.^26–31 Nowadays, this photophysical phenomenon is applied in organic electroluminescence (EL), and triplet exciton-utilized OLEDs with high efficiency and various colors have been successfully obtained.^14,32–36

Fig. 3 Photophysical mechanism of hRISC.

Sulfur lies in the second row of the periodic table and belongs to the oxygen family, with an outermost electron structure of 3s²3p⁴ and an empty d orbital (Fig. 4). A sulfur-containing molecule is a kind of sparkling structure in chemistry, which is put into a plethora of applications ranging from medicine to optics and nanotechnology to separation science.³⁷ For applications in OLEDs, sulphur-fused heterocyclic derivatives represent one of the most investigated heterocyclic compounds. Sulfoxide or sulfone with electron-deficient characteristics can be easily formed by oxidation of sulphur, which is beneficial to electron injection and transportation. These sulphur-fused heterocyclic derivatives are commonly used as host materials or emissive layers in OLEDs.^38–40 It is also easy to substitute sulphur atoms with selenium or tellurium, heavy atoms in the same main group, through typical replacement reactions, which greatly reduces the LUMO energy level and red-shifts the emission spectra. Compared with O atoms, S atoms in organic molecules are more likely to lead to a distorted spatial geometry, which is attributed to its larger atomic radius (extra electron orbitals).⁴¹

Fig. 4 Electron structure, type of hybridization and bonding in S atoms.

Research studies of efficient sulphur-containing chromophores have been carried out for a long time, and reports of these emitters are mainly based on dibenzothiophene (DBT), diphenyl sulfone (DPS), thiadiazole (BZ/NZ), and phenothiazine (PTZ). Herein, a review of the recent progress in the above heterocyclic emitters is presented. The fundamental principles regarding molecular design are discussed, and representative reported TADF and hot exciton sulphur-based luminogens are summarized and analyzed, along with their structure–property relationships and OLED applications. By summarizing these efficient emitters with sulphur-containing chromophores, we expect to provide serviceable design strategies to deliver more efficient sulphur-containing compounds with high OLED performance.

2. High-efficiency emitters based on S-containing chromophores

In this paper, emitters based on the following four derivatives containing S chromophores are discussed: DBT with a rigid structure, DPS with an electron-withdrawing capacity, BZ/NZ with a flat structure and electron-withdrawing capacity, and PTZ with multiple distorted conformations caused by the large diameter of S atoms.

2.1 DBT based emitters

DBT possesses a rigid structure. The sulphur atom in DBT undergoes sp³ unequal hybridization with two lone pairs of electrons, which usually serves as an electron-rich donor. In the early research of OLEDs, DBT is often used in polymer light-emitting diodes (PLEDs) owing to its good hole-transporting properties.^42–44 Recently, DBT has also been applied to synthesize TADF materials.

In 2017, Tang and co-workers reported a TADF-AIE (aggregation-induced emission)-ML (mechanoluminescence) emitter, DBT-BZ-DMAC (Fig. 5).⁴⁵DBT-BZ-DMAC had an unsymmetrical D–A–D′ structure, where benzoyl (BZ) served as an electron acceptor, and 9,9-dimethyl-9,10-dihydroacridine (DMAC) and DBT functioned as electron donors (D and D′, respectively). The DMAC and BZ formed a twisted conformation to realize AIE and TADF properties, while the planar DBT helped to increase the charge-transporting ability of the luminogen. The crystal analysis of DBT-BZ-DMAC showed that the BZ-DMAC segments with charge transfer (CT) properties form hydrogen bonds between molecules to obtain a high photoluminescence (PL) efficiency and a small red-shift in the solid state (Fig. 6). The PL efficiency of the 4,4′-di(9H-carbazol-9-yl)-1,1′-biphenyl (CBP, Fig. 20) doped film is 65.9%, lower than the 80.2% PL efficiency of the DBT-BZ-DMAC pure film, indicating that DBT-BZ-DMAC did not need doping to prevent emission quenching due to its AIE nature. Doped and non-doped OLEDs were fabricated by adopting DBT-BZ-DMAC as the emitter. The doped OLEDs with a low doping concentration afforded superb peak EL efficiencies, but the efficiency roll-off remained large, like most doped OLEDs with TADF emitters in the literature. With the increase of doping concentration, the peak efficiency values were decreased but the efficiency roll-off apparently became smaller. The non-doped OLEDs provided excellent peak EL efficiencies of 43.3 cd A⁻¹, 35.7 lm W⁻¹, and 14.2%, with an extremely small current efficiency (CE) roll-off of 0.46%, and an external CE roll-off approaching null from peak values to those at 1000 cd m⁻², enabling non-doped OLEDs to outperform doped OLEDs at high luminance. In doped OLEDs of DBT-BZ-DMAC show high peak current and external quantum efficiencies of up to 51.7 cd A⁻¹ and 17.9%, respectively, but the efficiency roll-off is large at high luminance. In the same year, Wang et al. calculated the quantum mechanics (QM) and molecular mechanics (MM) of DBT-BZ-DMAC and analyzed the XRD of its crystal and powder.⁴⁶ The results indicated that the non-radiative decay rate of DBT-BZ-DMAC in the solid phase was significantly decreased due to the suppression of the rotation of the DMAC and DBT units, while the radiative rate was greatly increased owing to the enhancement of the transition dipole moment. Both the intersystem crossing (ISC) and RISC processes happened between the S₁ and the lowest degenerate T₁ and T₂ states (Fig. 7). In addition, the CT rate was studied using the Marcus theory and the intrinsic charge mobility was calculated by performing the kinetic Monte Carlo method. The results showed that the DBT-BZ-DMAC crystal was a p-type semiconductor with a hole mobility of 0.14 cm² V⁻¹ s⁻¹ at room temperature.

Fig. 5 Molecular structures of the materials used in Section 2.1.

Fig. 6 (A and B) Packing patterns of DBT-BZ-DMAC in crystals. (C) Mechanoluminescence photo of the as-prepared DBT-BZ-DMAC crystalline powders, upon scratching under ambient conditions without UV illumination. Reproduced with permission from ref. 45. Copyright 2017 John Wiley & Sons.

Fig. 7 Adiabatic excitation energies for DBT-BZ-DMAC in the gas (a) and solid phases (b), respectively.

In 2017, Tang et al. prepared two molecules, DBT-BZ-PXZ and DBT-BZ-PTZ, in which the planar DBT was used as the invariable donor to increase the charge-transporting ability of the materials. Phenoxazine (PXZ) and PTZ showed good electron-donating and hole-transporting abilities and promoted the separation of the HOMO and LUMO with a BZ acceptor.⁴⁷ Both OLED materials based on the two molecules exhibited TADF-AIE properties. For example, for X-ray crystallography of DBT-BZ-PXZ, the dihedral angle between the donor PXZ and benzene reached 66° (Fig. 8), which facilitated the HOMO/LUMO separation and made the molecules loosely stack in the crystal lattice. There was no tight π–π stacking between DBT and BZ fragments. Multiple C–H⋯π and C=O⋯H hydrogen bonds were also observed. The non-radiative decay was inhibited by various intermolecular interactions and the luminescence efficiency of the concentrated state was improved. The maximum EQEs of the doped device based on DBT-BZ-PXZ and DBT-BZ-PTZ reached 15.1% and 19.2%, respectively. It is worth mentioning that the maximum EQE of non-doped devices based on DBT-BZ-PTZ was 9.7%, and the maximum EQE of 8.5% was still maintained under the brightness of 1000 cd m⁻². This work showed that the roll-off of undoped devices based on TADF-AIE materials could be significantly suppressed at high brightness.

Fig. 8 Crystal structure (CCDC 1455106) and packing pattern of DBT-BZ-PXZ in crystals. Reproduced with permission from ref. 47. Copyright 2017 American Chemical Society.

The manipulation of the emission properties of deep-blue emitters with TADF features through molecular design is challenging. In 2019, Dias et al. synthesized three TADF molecules, DAc-DBT, DtBuAc-DBT and DOMeAc-DBT, applying DBT as an unconventional acceptor to fulfill this goal.⁴⁸ The design of DAc-DBT derivatives was inspired by a previously synthesized molecule, DPTZ-DBT,⁴⁹ which exhibited excellent room temperature phosphorescence (RTP) emission. In the tert-butyl substituted compound the low energy triplet was localized on the acceptor unit (Fig. 9), with the RISC mechanism (k_RISC = 0.17 × 10⁵ s⁻¹) likely involving the mixture of CT and locally excited (LE) triplet states that were varied by less than 0.09 eV. An optimized OLED based on the tBu-compound presented a maximum EQE of 10.5% and a deep-blue emission with CIE coordinates of (0.133, 0.129).

Fig. 9 (a) Natural transition orbitals (NTO) corresponding to the first triplet transitions. (b) Singlet and triplet energy diagrams of DAc-DBT, DtBuAc-DBT and DOMeAc-DBT along with the energy levels of the individual units (TD-DFT rBMK/6-31G(d)).

2.2 DPS-based emitters

DPS is widely used in blue light emitting materials with a D–A structure owing to its relatively shallow LUMO energy level and distorted spatial conformation. In recent years, DPS has been acting as a classical acceptor for blue TADF materials with excellent device performance.^50–52

In 2012, Adachi's group published a paper on the synthesis of three blue TADF molecules DPA-DPS, TDPA-DPS and TDTC-DPS (Fig. 11) using DPS as the acceptor, and diphenylamine (DPA), di-tert-butyldiphenylamine (TDPA) and di-tert-butylcarbazole (TDTC) as the respective donors.⁵³ All three compounds exhibited broad and structureless emission bands with maxima between 402–419 nm, which could be ascribed to the intramolecular CT transition because of their dipolar nature. Their ΔE_ST values were calculated to be slightly raised for the ¹CT state, and considerably raised for the ³ππ* state, 0.54, 0.45, and 0.32 eV, respectively. The introduction of tert-butyl groups on the DPA unit enhanced its electron donating ability, and consequently lowered the CT energy and the ΔE_ST. On the other hand, the replacement of a DPA unit with a carbazole (Cz) unit slightly raised the ¹CT state, and considerably raised the ³ππ* state, resulting in a further decrease in its ΔE_ST. TDTC-DPS based OLEDs achieved the best performance with a maximum EQE of 9.9% and EL peaking at 423 nm (Fig. 10). This was a very impressive device performance at the time, and opened a relatively simple and intuitive path for the manufacture of efficient blue TADF materials.

Fig. 10 (a) The EQE–current density characteristics of the OLEDs based on (1) DPA-DPS, (2) TDPA-DPS and (3) TDTC-DPS. (b) The EL and PL spectra of 10 wt% TDTC-DPS doped in a DPEPO layer. Inset: The CIE coordinates of the EL spectrum of a TDTC-DPS-based device. Reproduced with permission from ref. 53. Copyright 2012 American Chemical Society.

Fig. 11 Molecular structures of the materials containing DPS units.

Subsequently, on the basis of TDTC-DPS, Adachi et al. replaced the tert-butyl group with a methoxyl group to produce a new compound DMOC-DPS (Fig. 11).⁵⁴DMOC-DPS exhibited a much shorter excited-state lifetime in both aromatic solution and organic thin film, because the change of the substituent on the donor affected the S₁ and T₁ excited states in different ways, decreasing the energy gap between the S₁ and T₁ states. DMOC-DPS possessed a smaller ΔE_ST (0.24 eV) relative to that of DTC-DPS in organic thin films, because the change of the substituent on the donor affected the S₁ and T₁ excited states in different ways, decreasing the energy gap between the S₁ and T₁ states. DMOC-DPS possessed a smaller ΔE_ST (0.24 eV) relative to that of DTC-DPS in toluene (0.32 eV). OLEDs based on DMOC-DPS achieved a maximum EQE of 14.5% with an EL peak at 460 nm and CIE coordinates of (0.16, 0.16), which was the highest level of blue OLEDs at the time. In addition, it is worth noting that the roll-off of DMOC-DPS was reduced and there was still a 9.0% EQE at 100 cd m⁻². This work successfully provided an effective path for the design of low roll-off OLED materials.

Adachi and co-workers continued to explore the DPS-based blue TADF material. In 2014, they reported three TADF molecules, PPZ-DPS, PXZ-DPS, and DMAC-DPS (Fig. 11).⁵⁰ A relatively large torsion angle of 89° was obtained between DMAC and DPS. Such a large degree of distortion resulted in a small overlap between the HOMO and LUMO in DMAC-DPS, resulting in a ΔE_ST less than 0.1 eV. The non-doped device based on DMAC-DPS eventually achieved the CIE coordinates of (0.16, 0.20), an EL emission peak at 470 nm and a surprisingly high maximum EQE of 19.5%, representing one of the best device performances so far. The ΔE_ST was small enough to significantly suppress the roll-off, and the EQE still remained at 16.0% at 1000 cd m⁻² brightness. This work provided valuable guidance for achieving efficient blue TADF materials.

In 2015, Adachi et al. further optimized the device structure of DMAC-DPS and prepared high-efficiency simple non-doped devices that could be as efficient as the best doped OLEDs.⁵⁵ A blue-emitting undoped OLED employing the DMAC-DPS derivative achieved an EQE of 19.5% at a luminescence of 100 cd m⁻² (Fig. 12). It was found that a relatively large Stokes’ shift and weak π–π stacking interactions were responsible for the concentration insensitive properties of the TADF emitter. This was another big advantage of TADF-based OLEDs compared with PHOLEDs.

Fig. 12 (a) Structures of undoped OLEDs containing DMAC-BP or DMAC-DPS and the chemical structures of the hole-transport and electron-transport materials. (b) Electroluminescence spectra of DMAC-BP-based OLEDs at 1000 cd m⁻². (c) The luminance–current density–voltage characteristics of DMAC-BP-based OLEDs. (d) The EQE–current density characteristics of DMAC-BP-based OLEDs. (e) Electroluminescence spectra of DMAC-DPS based OLEDs at 1000 cd m⁻². (f) The luminance–current density–voltage characteristics of DMAC-DPS based OLEDs. (g) The EQE–current density characteristics of DMAC-DPS based OLEDs. Reproduced with permission from ref. 55. Copyright 2012 John Wiley & Sons.

In 2019, Yang et al. reported two TADF-AIE molecules mono-DMACDPS (DMAC-1DPS) and Me-DMAC-DPS (Fig. 11) with the goal of designing multifunctional organic emitters.⁵⁶ mono-DMACDPS and Me-DMACDPS exhibited typical AIE, TADF, and room temperature phosphorescence (RTP) properties but different ML behaviors. Crystal structure analysis revealed that a large dipole moment and multiple intermolecular interactions with a tight packing mode endowed mono-DMACDPS with strong ML (Fig. 13). The introduction of a methyl group reduced the number and intensity of molecular interactions. Me-DMAC-DPS exhibited strong π–π interactions, which was seriously detrimental to ML emission. The separated monomer and dimer in the crystal led to typical TADF and RTP properties, respectively. mono-DMACDPS was the first example realizing TADF, RTP, AIE, and ML simultaneously. In addition, it is worth noting that the ΔE_ST values of DMAC-1DPS and Me-DMAC-DPS were as low as 0.014 eV and 0.015 eV, respectively, indicating the possible existence of efficient RISC channels with high triplet exciton utilization.

Fig. 13 (a) The ML image of DMAC-1DPS. (b) ML spectrum in the crystal. (c and d) Crystal structures and intermolecular interactions for DMAC-1DPS and Me-DMACDPS, respectively. (e) Image of the DMAC-1DPS crystal with a 365 nm lamp on and off. Reproduced with permission from ref. 56. Copyright 2012 John Wiley & Sons.

Yang et al. further synthesized a TADF molecule, SBA-2DPS (Fig. 11), using DPS as the electron acceptor and spirodiacridine (SBA) as the electron donor.⁵¹ The spiro-linked double D–A molecular architecture was introduced to selectively improve the horizontal emitting dipole orientation. Compared to the prototypical DMAC-1DPS, the shape of SBA-2DPS was elongated without expanding the degree of whole conjugation owing to the s-spacer linkage (Fig. 14). Consequently, a high horizontal dipole ratio of 87% was achieved for SBA-2DPS. The blue OLEDs based on SBA-2DPS showed a higher maximum EQE of 25.5% than DMAC-1DPA, with CIE coordinates of (0.15, 0.20). The effect of horizontal emission dipole and optical microcavity was considered to be an important factor to improve the EQE.

Fig. 14 (a) Molecular structures of DMAC-1DPS and SBA-2DPS. (b) Optimized ground state structures. (c) The HOMO (blue) and LUMO (red) distributions with optimized ground state structures. (d) Isodensity surface of the designed molecules with r = 0.001 e bohr⁻³.

Subsequently, Su's team also constructed a similar TADF emitter PX-SBA (Fig. 11) by utilizing SBA as the central donor, and PX, the two phenyl groups of DPS bridged by ether bonds, as the acceptor. Benefiting from the horizontal orientation and TADF feature, PX-SBA exhibited satisfactory electroluminescence performance overall. Pure blue emission with CIE coordinates of (0.16, 0.13) and an ideal EQE value of 20.8% were obtained in the doped devices.⁵⁷

In 2019, Chi's group constructed two D–A–D* type TADF molecules, 4Cz-DPS and 2Cz-DPS (Fig. 11), by introducing a Cz donor unit at the ortho-position, in which the donor and acceptor groups were spatially in close proximity to guarantee the existence of intramolecular electrostatic attraction and through-space charge transfer.⁵⁸ Therefore two charge transfer pathways existed, namely bond charge transfer and space charge transfer (Fig. 15), which could enhance the radiation attenuation intensity of the emitter. The emission peak and photoluminescene quantum yield (PLQY) of pure films based on 4Cz-DPS and 2Cz-DPS were 530 nm and 74.7%, and 520 nm and 65.3%, respectively. In addition, 2Cz-DPS also exhibited AIE properties, which were beneficial for the fabrication of non-doped devices. The non-doped devices based on 4Cz-DPS and 2Cz-DPS eventually achieved a maximum EQE of 20.7% and 28.7%, which represented one of the best performances of non-doped TADF-OLEDs.

Fig. 15 (a) Single crystal and (b) reduced density gradient (RDG) isosurface map with an isovalue of 0.5 for 2Cz-DPS. (c and d) The functions of RDG and Sign(λ₂)_ρ for 2Cz-DPS and 4Cz-DPS, respectively. (e) HOMO and LUMO distribution of 2Cz-DPS, showing the overlapping integral extents I_H/L.

Although DPS had been used widely as the an acceptor for blue TADF, the ΔE_ST of some molecules mentioned above was not small enough. Kido et al. proposed a solution in 2015, that is, to introduce another phenylsulfone into the DPS acceptor core to regulate the length of molecular π-conjugation and the electronic properties of the donor and acceptor, expecting to reduce ΔE_ST. TADF materials, DTC-pBPSB and DTC-mBPSB adopting (1,4-bis(benzenesulfonyl)benzene) and 1,3-bis(benzenesulfonyl)benzene as acceptors, were produced respectively.⁵⁹ It should be pointed out that pBPSB-based compounds possessed narrower energy band gaps and lower singlet and triplet energy levels compared with mBPSB-based isomers, due to the shorter effective conjugation and weaker electron-absorbing capacity. The ΔE_ST values of DTC-pBPSB and DTC-mBPSB were estimated to be 0.05 and 0.24 eV, respectively. DTC-mBPSB-based OLEDs exhibited deep blue CIE coordinates of (0.15, 0.08) with a maximum EQE of 5.5%, while DTC-pBPSB OLEDs exhibited sky blue CIE coordinates of (0.18, 0.19) with a maximum EQE of 11.7%. The small ΔE_ST and good device performance demonstrated the feasibility of the BPSB design and increased the variety of TADF acceptors based on phenylsulfone.

Kido and co-workers further put forward a strategy to achieve fluorescent materials simultaneously possessing a high-energy singlet CT state and triplet states, as well as an efficient radiative process via inserting a twisted core into the molecular backbone. In this work, the electron-donor DTC was replaced by 9,9-dimethyl-9,10-dihydroacridine (ACR) to generate ACR-pBPSB.

ACR-BPSBP (Fig. 11) was also prepared for studying the subtle influences of the inner-connection type on the molecular CT interactions.⁵² According to DFT calculations (Fig. 16), the two benzene rings in the BPSBP core were almost orthogonal to each other, which made the two sulfonyl groups of ACR-BPSBP show a weak electron deficiency ability and led to an increased band gap, which eventually resulted in a shorter maximum emission wavelength. Besides, because of the high energy of the ³CT state, the difference between E⁰ (³CT) and E⁰ (³LE) of ACR-BPSBP was relatively bigger than that of ACR-pBPSB, which suggested a relatively inefficient TADF process for ACR-BPSBP, and was proven by the following test of the lifetime of the delayed fluorescence (τ_DF). For devices based on ACR-pBPSB and ACR-BPSBP, the maximum EL emission was observed at 468 nm and 502 nm, respectively. The CE of the device based on ACR-pBPSB was 37.6 cd A⁻¹ and the maximum EQE was 13.5%. The CE of the device based on ACR-BPSBP was 38.7 cd A⁻¹ and the EQE was 24.6%, which was almost doubled. The final results showed that the type of internal connection could influence the molecular CT interaction.

Fig. 16 (a) Calculated distributions and energy levels of the LUMO (top) and HOMO (bottom) and the molecular surface electrostatic potential (middle) in S₀ geometries (simulated at the B3LYP/6-31G* level, blue and red for positive and negative electrostatic potential, respectively). (b) EQE versus current density characteristics of the electroluminescence devices based on ACR-BPSBP and ACR-pBPSB. (c) The schematic diagram of influence came from the energy gap between two triplet states (³CE and ³LE) on the whole TADF process; S₀, ground state; IC, internal conversion; RIC, reverse internal conversion; ISC, intersystem crossing; RISC, reverse intersystem crossing. (d) EL spectra of the electroluminescence devices based on ACR-BPSBP and ACR-pBPSB measured at 5 mA cm⁻². Reproduced with permission from ref. 52. Copyright 2017 John Wiley & Sons.

In fact, sulfone not only can be utilized to construct TADF emitters as the suitable acceptor, but also acts as a medium-strength acceptor for hot exciton materials. In 2015, Lu et al. reported a hot exciton material PMSO with a D–A–D structure by selecting PPI as a weak donor and DPS as a weak acceptor.⁶⁰ A highly mixed or hybrid local and charge transfer (HLCT) excited state would simultaneously achieve a large fraction of singlet formation and a high PLQY. The PMSO-based doped device displayed deep blue electroluminescence with emission peak at 445 nm and CIE coordinates of (0.152, 0.077). The maximum brightness, EQE, and CE were 15 599 cd m⁻², 6.80%, 4.64 cd A⁻¹, respectively. This work demonstrated the feasibility of HLCT molecular design in achieving efficient deep blue electroluminescence. Lu et al. reported deep-blue hybridized local and charge-transfer (HLCT) material compounds, DP-TXO₂ and P-TXO₂ (Fig. 17), in which 9,9-dimethylthioxanthene-S,S-dioxide (TXO₂) with a LUMO energy level of −2.33 eV was selected as the acceptor, and pyrene was used as the donor to ensure deep-blue emission.⁶¹ To alleviate the intermolecular interactions between the pyrene subunit of the objective compounds and the 4,4′,4′′-tris(carbazol-9-yl)triphenylamine (TCTA, Fig. 20) host, a bulky 1,1′:3,1′′-terphenyl (TP) was grafted at the pyrene moiety of P-TXO₂ to render TPP-TXO₂. As a result, TPP-TXO₂ achieved the best device performance with CIE coordinates of (0.152, 0.065), EQE_max of 10.5%, as well as a high brightness of 10 480 cd m⁻². The maximum EUE of this device was even close to 100%.

Fig. 17 Molecular structures of the materials containing derivatives of DPS.

Modification of the molecular structure of DMAC-DPS also could afford the TADF emitter in the deep blue region. Lee et al. reported the deep-blue light emitting molecule DMTDAc (Fig. 17) using a rigid 9,9-dimethyl-9H-thioxanthene-10,10-dioxide (DMTD) acceptor and an acridine donor, aiming to obtain high-efficiency deep blue TADF and a smaller FWHM than DMAC-DPS.⁶² The inter-locking two phenyl units of DPS could reduce the molecular motion and shifted the emission color to the deep blue region. Therefore, the broad emission spectrum of the common TADF emitter became narrower by engineering the molecular structure to restrict the molecular motion. OLEDs based on DMTDAc achieved the CIE coordinates of (0.15, 0.13), a FWHM of 65 nm and an EQE close to 20%.

Table 1 Summary of the device performances of emitters containing DBT units

Emitter	Emitting layer	λ EL [nm]	CE^b [cd A^−1]	PE^c [lm W^−1]	EQE^d [%]	CIE 1931 [x, y]	Ref.
a EL peak of the device. b The maximum value and value at 1000 cd m^−2 of current efficiency. c The maximum value and value at 1000 cd m^−2 of power efficiency. d The maximum value and value at 1000 cd m^−2 of EQE.
DBT-BZ-DMAC	6 wt% emitter:CBP	508	51.7/31.1	50.7/17.2	17.9/10.9	0.234, 0.510	45
	emitter	516	43.3/43.1	35.7/33.1	14.2/14.2	0.261, 0.550
DBT-BZ-PXZ	6 wt% emitter:CBP	528	60.6/48.6	59.2/21.8	19.2/15.4	0.340, 0.576	47
	emitter	557	26.6/19.6	27.9/11.3	9.2/6.8	0.434, 0.542
DBT-BZ-PTZ	10 wt% emitter:CBP	538	46.0/29.5	43.3/13.6	15.1/9.7	0.370, 0.563	47
	emitter	563	26.5/23.5	29.1/15.4	9.7/8.5	0.448, 0.531
DtBuAc-DBT	10 wt% emitter:DPEPO	455	13.9/-	9.7/-	10.5/-	0.133, 0.129	48

---

Table 2 Summary of the device performances of emitters containing DPS units

Emitter	Emitting layer	λ EL [nm]	CE^b [cd A^−1]	PE^c [lm W^−1]	EQE^d [%]	CIE 1931 [x, y]	Ref.
a EL peak of the device. b The maximum value and value at 1000 cd m^−2 of current efficiency. c The maximum value and value at 1000 cd m^−2 of power efficiency. d The maximum value and value at 1000 cd m^−2 of EQE.
DPA-DPS	10 wt% emitter:DPEPO	421	—	—	2.9	—	53
TDPA-DPS	10 wt% emitter:DPEPO	430	—	—	5.6	—
TDTC-DPS	10 wt% emitter:DPEPO	423	—	—	9.9	0.15, 0.07
DMOC-DPS	10 wt% emitter:DPEPO	460	24.0/—	—	14.5/3.7	0.16, 0.16	54
PXZ-DPS	Emitter:CBP	—	—	—	17.5/15.5	—	50
DMAC-DPS	Emitter:DPEPO	470	—	—	19.5/16.0	0.16, 0.20
	Emitter	480	—	—	19.5/14.6	0.16,0.29	55
DMAC-1DPS	30 wt% emitter:DPEPO	477	31.0/20.9	30.5/10.9	17.4/11.6	0.16,0.26	56
SBA-DPS	30 wt% emitter:DPEPO	467	38.1/23.5	37.4/12.3	25.5/15.6	0.15, 0.20	51
PX-SBA	20 wt% emitter:DPEPO	448	24.4/8.3	24.0/3.7	20.8/7.0	0.16, 0.15	57
4Cz-DPS	Emitter	524	61.2/—	38.4/—	20.7/—	—	58
2Cz-DPS	Emitter	518	82.3/—	51.8/—	28.7/—	—
DTC-mBPSB	10 wt% emitter:DPEPO	≈444	4.4/—	—	5.5/—	0.15, 0.08	59
DTC-pBPSB	10 wt% emitter:DPEPO	≈474	19.4/—	—	11.7/—	0.18, 0.19
ACR-pBPSB	20 wt% emitter:mCP	502	37.6/—	—	13.5/—	0.16, 0.21	52
ACR-BPSBP	20 wt% emitter:DPEPO	468	38.7/—	—	24.6/—	0.27, 0.48
PMSO	10 wt% emitter:CBP	445	4.64/—	4.00/—	6.80/5.64	0.152, 0.07	60
	Emitter	465	7.31/—	6.23/—	4.95/4.86	0.157, 0.206
DMTDAc	30 wt% emitter:UGH3/TSPO1	451	22.6/—	23.2/—	19.8/—	0.15, 0.13	62
P-TXO2	12 wt% emitter:TCTA/12 wt% emitter:26DCzPPy	452	3.0/—	—	4.0/3.4	0.106, 0.085	61
DP-TXO2	12 wt% emitter:TCTA/12 wt% emitter:26DczPPy	462	4.6/—	—	4.0/3.5	0.154, 0.098
TPP-TXO2	12 wt% emitter:TCTA/12 wt% emitter:26DczPPy	443	11.1/—	—	10.5/4.6	0.152, 0.065
CZ-MPS	10 wt% emitter:TCTA/10 wt% emitter:CzSi	389	3.4/—	—	9.3/	0.175, 0.065	63
DPAC-DBTDO	15 wt% emitter:DPEPO	500	34.2/—	26.3/—	13.1/5.7	0.252, 0.455	65
TXO-TPA	5 ± 1 wt% emitter:mCP	552	43.3/—	47.7/—	18.5/—	0.45, 0.53	66
TXO-PhCz	5 ± 1 wt% emitter:mCP	—	76.0/—	70.0/—	21.5/—	0.31, 0.56
TXO-PhCz3	10 wt% emitter:mCP	532	45.3/6.8	35.6/2.2	14.0/2.1	0.40, 0.54	68
TXO-PhCz4	6 wt% emitter:PPO21	540	49.4/26.8	47.714.7	16.3/9.0	0.38, 0.55
2,3-TXO-PhCz	15 wt% emitter:CBP	—	37.2/—	27.8/—	11.9/—	0.42, 0.55	69
2,6-TXO-PhCz	8 wt% emitter:CBP	—	72.3/—	64.9/—	23.2/—	0.40, 0.55
2,7-TXO-PhCz	15 wt% emitter:CBP	—	74.6/—	68.9/—	24.4/—	0.39, 0.55
3,6-TXO-PhCz	8 wt% emitter:CBP	—	54.3/—	48.8/—	18.1/—	0.43, 0.53
PXZ2PTO	Emitter	504	44.9/—	32.0/—	16.4/—	0.27, 0.50	70
TXADO-spiro-DMACF	Emitter	444	5.3/—	5.9/—	5.3/—	0.16, 0.09	71
DMA-ThX	Emitter:DPEPO	462	18.2/—	—	18.4/—	0.14, 0.14	72

---

Table 3 Summary of the device performances of emitters containing BZ units

Emitter	Emitting layer	λ EL [nm]	CE^b [cd A^−1]	PE^c [lm W^−1]	EQE^d [%]	CIE 1931 [x, y]	Ref.
a EL peak of the device. b The maximum value and value at 1000 cd m^−2 of current efficiency. c The maximum value and value at 1000 cd m^−2 of power efficiency. d The maximum value and value at 1000 cd m^−2 of EQE.
TPA-BZP	Emitter	588	8.84/7.92	7.18/4.64	3.80/3.41	0.55, 0.45	74
CzP-BZP	Emitter	538	23.99/13.18	16.38/8.82	6.95/3.81	0.34, 0.60	75
BTH-DMF	30 wt% emitter:SPPO1	550	31.02/16.95	27.85/8.43	9.13/4.94	0.41, 0.56	76
2F-BTH-DMF	30 wt% emitter:SPPO1	515	26.62/12.57	26.14/5.25	8.52/4.02	0.29, 0.54
o-BTH-DMF	30 wt% emitter:SPPO1	454	1.73/0.79	1.16/0.40	1.51/0.69	0.15, 0.13
PTZ-BZP	Emitter	692	—	—	1.54/—	0.69, 0.30	36
BTDF-TTPA	Emitter	690	0.11/—	0.09/—	0.83/—	0.70, 0.30	77
	1 wt% emitter:CBP	630	5.15/—	4.27/—	5.75/—	0.61, 0.34
BTDF-TtTPA	Emitter	690	0.19/—	0.16/—	1.44/—	0.71, 0.29
	1 wt% emitter:CBP	642	2.98/—	2.60/—	4.94/—	0.63, 0.33
TPABTPA	Emitter	608	17.7/—	12.4/—	11.1/—	0.59, 0.40	78
	Emitter:CBP	608			18.1	0.58, 0.41
TPABCHO	Emitter	640	4.7/—	2.5/—	5.0/—	0.64, 0.36
	Emitter:CBP	590			8.9	0.54, 0.45

---

Table 4 Summary of the device performances of emitters containing NZ units

Emitter	Emitting layer	λ EL [nm]	CE^b [cd A^−1]	PE^c [lm W^−1]	EQE^d [%]	CIE 1931 [x, y]	Ref.
a EL peak of the device. b The maximum value and value at 1000 cd m^−2 of current efficiency. c The maximum value and value at 1000 cd m^−2 of power efficiency. d The maximum value and value at 1000 cd m^−2 of EQE.
TPA-NZP	Emitter	664	1.00/—	0.77/—	2.8/—	0.67, 0.32	79
TPA-NZC	Emitter	702	—	—	1.2/—	0.685, 0.299	80
	6 wt% emitter:CBP	656	—	—	3.2/—	0.657, 0.335
NZ2TPA	Emitter	696	—	—	3.9/2.8	0.70, 0.30	81
NZ2AC	Emitter	663	2.6/—	2.0/—	2.8/2.0	—	82
	8 wt% emitter:CBP	612	11.2/—	11.3/—	6.2/2.4	—
PXZ-3-NZP	Emitter	738	—	—	0.82/—	0.703, 0.289	83
	7wt% emitter:MADN	672	—	—	2.03/—	0.686, 0.306
PXZ-10-NZP	Emitter	680	—	—	0.077/—	0.687, 0.304
	7wt% emitter:MADN	640	—	—	0.60/—	0.644, 0.354
TNZPPI	Emitter	686	0.40/—	0.47/—	2.48/—	0.69, 0.30	73
	10 wt% emitter:CBP	648	3.45/—	2.63/—	6.83/—	0.67, 0.33
TNZtPPI	Emitter	686	0.30/—	0.33/—	1.60/—	0.69, 0.31
	5 wt% emitter:CBP	642	4.01/—	3.31/—	6.52/—	0.64, 0.35
TNZ2tPPI	Emitter	678	0.31/—	0.32/—	1.48/—	0.69, 0.31
	10 wt% emitter:CBP	654	1.94/—	1.59/—	4.83/—	0.67, 0.33

---

Table 5 Summary of the device performances of emitters containing PTZ units

Emitter	Emitting layer	λ EL [nm]	CE^b [cd A^−1]	PE^c [lm W^−1]	EQE^d [%]	CIE 1931 [x, y]	Ref.
a EL peak of the device. b The maximum value and value at 1000 cd m^−2 of current efficiency. c The maximum value and value at 1000 cd m^−2 of power efficiency. d The maximum value and value at 1000 cd m^−2 of EQE.
PTZ-TRZ	2 wt% emitter:mCBP	532	—	—	10.8 ± 0.5/—	—	86
PTZ-DBPHZ	10 wt% emitter:CBP	613	19.6/—	—	16.8/—	—	93
CP-BP-PTZ	Emitter	554	46.1/38.4	55.7/30.2	15.3/12.7	0.42, 0.55	95
	10 wt% emitter:CBP	542	62.4/43.4	57.3/22.0	19.6/13.6	0.38, 0.57
PTZ-TTR	14.7 wt% emitter:CBP	448, 584	5.20/1.75	4.93/0.82	2.68/0.93	0.33, 0.33	98
PTZ-Ph-TTR	1.2 wt% emitter:CBP	456, 568	45.21/30.55	41.75/19.58	16.34/11.04	0.41, 0.47
PTZ-BP	4 wt% emitter:DCzDPy	440, 540	16.6/—	13.0/—	6.2/—	0.34, 0.46	100
	4 wt% emitter:CBP	—	15.5/—	—	5.44/—	—
PTZ-Mes2B	Emitter	540	62.88/—	—	19.66/17.31	0.37, 0.57	101
PTZ-ND	7.5 wt% emitter:mCP	528	40.9/34.3	31.4/22.0	13.0/11.0	0.30, 0.50	102
pipd-BZ-PTZ	6 wt% emitter:CBP	545	55.41/40.92	58.03/27.95	15.77/12.40	0.31, 0.57	103
	20 wt% emitter:CBP	550	38.70/37.80	31.05/26.99	12.03/11.76	0.42, 0.55

---

Subsequently, through interlocking of the DPS acceptor unit of a reported TADF emitter (CZ-PS)⁵³ by a dimethylmethylene bridge, CZ-MPS (Fig. 17), a UV-emissive TADF compound bearing a shallower LUMO energy level of −2.27 eV and a more rigid structure was achieved by Lu et al.⁶³ Compared with DPS, the two phenyl groups of MPS were partially interlocked and bridged by dimethyl methyl. Moreover, the rigidity of MPS inhibited the intermolecular motion, which was conducive to the narrowing of the PL spectrum. The λ_EL, CIE coordinates, maximum brightness and EQE_max were obtained to be 389 nm, (0.175, 0.065), 1031 cd m⁻² and 9.3%, respectively, for CZ-MPS doped OLED devices (Fig. 18). This work represented one of the highest efficiencies of UV-OLEDs.

Fig. 18 (a) EL spectrum of the CZ-MPS-based OLEDs at a current density of 100 mA cm⁻². (b) Current density–voltage–luminance (J–V–L) characteristics of the CZ-MPS-based OLEDs. Inset: EQE–current density curve of the CZ-MPS-based OLEDs. Reproduced with permission from ref. 63. Copyright 2020 John Wiley & Sons.

Similar to DMTD, dibenzo[B,D]thiophene-5,5-dioxide (DBTDO) can also be utilized as a good acceptor with a rigid structure. In 2013, Monkman et al. carried out a detailed theoretical study on TADF molecules containing linear and angular DBTDO groups, and demonstrated that the material structure, that is, angular versus linear structures, and the presence of a ³nπ* triplet state and CT states with very small exchange energy were required to fully explain the observation of TADF in ICT materials with relatively large singlet ¹CT-triplet ³ππ* energy splitting.⁶⁴ This TADF mechanism was shown to be nearly 100% efficient in converting the triplet to singlet states. The key role of the ³nπ* could also explain the recent results of Adachi and co-workers who demonstrated near 100% internal quantum efficiency in a device where the ³CT state was higher in energy than the triplet energy of the host material used which should efficiently quench the ³CT state. Both the ³nπ* and ³ππ* states would be lower in energy than the host so as to avoid quenching. In 2016, Lu et al. prepared a TADF molecule DPAC-DBTDO (Fig. 17) with a D–A–D structure by using DBTDO as the acceptor and 9,9-diphenyl-9,10-dihydroacridine (DPAC), which possessed large steric hindrance and a rather high triplet energy when substituted on the 9-site, as the donor.⁶⁵ The PL and transient fluorescence spectra showed that DPAC-DBTDO possessed a small ΔE_ST of 0.10 eV, which was favorable for an efficient RISC process. The doped OLEDs exhibited stable blue-green fluorescence with the CIE coordinates of (0.252, 0.455), a maximum CE of 34.2 cd A⁻¹, and an EQE_max of 13.1%.

In addition to DMTD and DBTDO, TADF based on other rigid phenyl sulfones also attracted much attention. In 2014, Wang's group synthesized two novel TX-based emitters with excellent TADF properties, TXO-TPA and TXO-PhCz (Fig. 17). Both of them had a typical D–A structure with 9-H-thioxanthen-9-one-10,10-dioxide (TXO) as the acceptor and triphenylamine (TPA)/N-phenylcarbazole (N-PhCz) as the donor.⁶⁶ The AIE characters of TXO-PhCZ and TXO-TPA were observed in water/acetonitrile mixtures, with PLQYs of 0.93 ± 0.02 and 0.36 ± 0.02 for films, respectively. The ΔE_ST were estimated to be 0.04 eV for TXO-TPA and 0.09 eV for TXO-PhCz. The maximum EQE value of TXO-TPA and TXO-PhCz reached 21.5% and 18.5%, respectively (Fig. 19). It was the record level for OLEDs based on the TADF mechanism at that time.

Fig. 19 (a) Current density–voltage–luminance characteristics of the OLEDs based on compounds TXO-TPA and TXO-PhCz. (b) The EQE–current density characteristics. (c) The EL spectra when operated at different voltages of the device incorporating TXO-TPA. (d) The EL spectra when operated at different voltages of the device incorporating TXO-PhCz. Reproduced with permission from ref. 66. Copyright 2014 John Wiley & Sons.

Fig. 20 Molecular structures of the materials used as the host in the emitting layers of the devices listed in Tables 1–5.

And then, in 2016, they further optimized the host material of the OLEDs and selected 1,3,5-tris(N-phenylbenzimidazole-2-yl)benzene (TPBI) as the doped host to adjust the interaction between the host and guest.⁶⁷ Strong interaction between TXO-PhCz and TPBI could be observed, leading to the lower singlet–triplet energy gap of 8.8 meV and non-monotonic increase of PLQYs (Φ_Totle, Φ_D, and Φ_P) with temperature. Finally, the performance of the TXO-PhCz-based device was further improved, achieving a maximum CE of 71.9 cd A⁻¹, a maximum power efficiency of 45.2 lm W⁻¹ and a maximum EQE of 23.2%. This work demonstrated the importance of choosing the right host for the EL performance.

In 2017, they further designed three D–A isomers based on TXO, TXO-PhCz1, TXO-PhCz3 and TXO-PhCz4 (Fig. 17) to study the effect of PhCz substituted at different positions.⁶⁸TXO-PHCz4 exhibited strong orbital coupling resulting in strong phosphorescence emission, which eventually caused the inverse temperature dependence of the transient PL decay, contrary to that of TXO-PhCz3 and other TADF emitters. TXO-PhCz4 exhibited a small ΔE_ST of 23 meV and a short decay time of 14 ms at room temperature, which were much smaller and shorter than those of TXO-PhCz3 and made the efficiency roll-off of the device very slight. The maximum CE, PE and EQE of the devices containing TXO-PhCz4 were 45.6 cd A⁻¹, 35.8 lm W⁻¹ and 15.4%, respectively.

In 2018, they further reported four D–A–D isomers, 2,3-TXO-PhCz, 2,6-TXO-PhCz, 2,7-TXO-PhCz, and 3,6-TXO-PhCz (Fig. 17).⁶⁹ The substitution positions of the two PhCz units significantly impacted the photophysical properties of the isomers, especially for the ΔE_ST and oscillator strength (f) (Fig. 21). The four emitters possessed a small ΔE_ST of 0.01–0.24 eV; the corresponding f values were 0.064, 0.107, 0.026, and 0.134. This result indicated that 3-position substitution contributed to the increase of f and thus improved the PLQY. The device based on 2,7-TXO-PhCz achieved the best device performance: 74.6 cd A⁻¹ for CE, 68.9 lm W⁻¹ for PE, and 24.4% for EQE. This work showed that the ΔE_ST and f played a synergistic role in OLED performance.

Fig. 21 (a–d) HOMO, HOMO−1, and LUMO electronic distributions, energy level, and bandgaps (E_g), the lowest singlet (S₁), and triplet (T₁) and oscillator strengths of S₁ (f) for 2,3-TXO-PhCz, 2,6-TXO-PhCz, 2,7-TXO-PhCz, and 3,6-TXO-PhCz calculated using the TD-DFT at the B3LYP/6-31G** level, respectively. Reproduced with permission from ref. 69. Copyright 2018 John Wiley & Sons.

TADF materials that can exhibit high efficiency in non-doped OLEDs are of significant importance. In 2018, Wang et al. reported an emitter PXZ2PTO (Fig. 17) which contained phenoxazine as the donor and 2PTO as the acceptor, exhibiting both AIE characteristics and TADF properties.⁷⁰PXZ2PTO exhibited a small ΔE_ST of 0.02 eV and a high PLQY of 68.75% in doped films and 61.54% in pure films. PXZ2PTO showed high efficiency in both undoped and doped devices. Green TADF non-doped OLEDs achieved a maximum EQE, CE and PE of 16.4%, 44.9 cd A⁻¹ and 32.0 lm W⁻¹, respectively, which was comparable to those of the doped OLEDs. These results indicated that PTZ-oxide could be a good acceptor for the design of aggregation induced delayed fluorescence (AIDF) materials.

Spiro-blocking is also a promising strategy to develop deep-blue TADF emitters capable of application in non-doped OLEDs. In 2019, Ding et al. reported a deep-blue TADF emitter TXADO-spiro-DMACF (Fig. 17) by integrating an appropriate blocking unit with the D−A−D-type TADF emitter via a spiro linkage.⁷¹ The intermolecular interactions were weakened to some degree benefiting from the characteristic perpendicular arrangement. As a result, TXADO-spiro-DMACF showed a very small bathochromic shift of 8 nm associated with a narrowed FWHM of 54 nm from solution to film. The corresponding non-doped device successfully achieved a bright deep-blue emission, revealing CIE coordinates of (0.16, 0.09) and a peak EQE of 5.3% (5.3 cd A⁻¹, 5.9 lm W⁻¹).

In 2019, Kim et al. synthesized a TADF material DMA-ThX (Fig. 17) using a simple process.⁷²DMA-ThX showed twisted moieties, and the bandgap was found to be sufficiently large to radiate deep-blue light with high-lying HOMO and LUMO energy levels of −5.77 and −2.37 eV, respectively. A small ΔE_ST of 0.07 eV for the TADF emission process was demonstrated by the extremely fast PL lifetimes of the DMA-ThX solution and films at room temperature and low temperature, which was in the order of 10⁻⁸ s. OLEDs reached a maximum EQE of 18.4% with CIE coordinates of (0.14, 0.14) when DMA-ThX was doped with bis[2-(diphenylphosphino)phenyl]ether oxide (DPEPO, Fig. 20), and the FWHM of the device was 65 nm.

2.3 BZ/NZ based emitters

As two main derivatives of thiazoles used in OLED emitters, BZ and NZ have been widely used in the construction of hot exciton materials.^36,73 The energy gap between the T₁ and T₂ states of BZ is large, and at the same time the energy gap between the S₁ and T₂ states is small, which meets the design concept of hot exciton materials.

In 2014, Ma and his colleagues reported a BZ-based molecule, TPA-BZP (Fig. 22), and further elaborated the emission mechanism of hot exciton materials through detailed characterization of TPA-BZP and its related molecules.⁷⁴ The non-doped OLED device of TPA-BZP achieved a maximum EQE of 3.8%. The maximum exciton utilization efficiency (EUE) was harvested up to 93%. Through the energy-level structure of the excited states, the hot exciton mode was proposed to be responsible for these large breakthroughs in exciton utilization efficiency.

Fig. 22 Molecular structures of the materials containing BZ units.

In 2016, Ma et al. further developed another molecule CzP-BZP (Fig. 22) on the basis of former TPA-BZP.⁷⁵ By switching the donor unit, that is, the stronger donor TPA was replaced by the weaker donor carbazobenzene (CzP), the relative distribution ratio of the LE state and CT state in CzP-BZP greatly changed (Fig. 23), and the PLQY of films based on CzP-BZP was increased by 30%. The EUE of 48% in CzP-BZP was also comparable to that of 42% in TPA-BZP. CzP-BZP exhibited an excellent EL performance in non-doped devices, achieving a maximum EQE of 6.95% and a maximum CE of 23.99 cd A⁻¹ with a CIE of (0.34, 0.60).

Fig. 23 Hybridization between LE and CT states at the geometry of the S₁ state of CzP-BZP and TPA-BZP, respectively. Reproduced with permission from ref. 75. Copyright 2016 American Chemical Society.

In 2018, Wang et al. designed and synthesized three D–A–D molecules 2F-BTH-DMF, BTH-DMF, and o-BTH-DMF (Fig. 22) applying BZ as the acceptor.⁷⁶ The 4,7-bis(9,9-dimethyl-9H-fluoren-2-yl) (DMF) unit was introduced due to its rigid and planar structure with weak electron-donating properties. The RISC process via the high-lying triplet state, also called the hot exciton mechanism, was adequately demonstrated by the steady and transient photophysical results combined with DFT calculations (Fig. 24). The PLQY of BTH-DMF and 2F-BTH-DMF was as high as 90.4% and 97.7%, respectively. The three emitters all showed high RISC rate constants k, and the k_RISC of BTH-DMF and 2F-BTH-DMF reached 10⁶ s⁻¹, which were higher than the 10⁵ s⁻¹ of o-BTH-DMF, making the oxygen quenching of o-BTH-DMF more obvious. The BTH-DMF device exhibited impressive CE, PE, and EQE up to 31.02 cd A⁻¹, 27.85 lm W⁻¹, and 9.13% with a maximum brightness of 25 400 cd m⁻², respectively.

Fig. 24 Theoretical calculation for the three compounds: natural transition orbital (NTO) pairs for the excited states of BTH-DMF (a), 2F-BTH-DMF (b), and o-BTH-DMF (c), respectively. (d) The energy diagram of the first ten singlet and triplet excited states of BTH-DMF, 2F-BTH-DMF, and o-BTH-DMF. Reproduced with permission from ref. 76. Copyright 2018 John Wiley & Sons.

As mentioned above, weak donor bonding with BZ could produce yellow-green fluorescence. Therefore, for these D–A structure molecules, a stronger donor unit could red-shift the butterfly phenothiazine PTZ donor.³⁶ The maximum emission peak was at 700 nm with a PLQY of 16% in thin films. The maximum EQE of undoped NIR OLEDs based on PTZ-BZP reached 1.54% and a low efficiency roll-off was observed, as well as a high radiative exciton ratio of 48%, which broke through the limit of 25% in conventional fluorescent OLEDs. According to DFT calculations, there existed a large energy level difference between the T energy levels, and the ΔE_ST of the S₁ and T₃ states was close to 0 (Fig. 25), which provided an efficient RISC channel. This work provided a new idea for the design of efficient NIR fluorescent molecules.

Fig. 25 (a) Natural transition orbitals for S₁, T₁, and T₃ states of PTZ-BZP. (b) The energy landscape for singlet and triplet excited states. (c) Model for exciton relaxation in the EL process. RISC: reverse intersystem crossing; IC(T): internal conversion between the triplet states; fl: fluorescence; ph: phosphorescence. (d) EQE–current–density characteristics of the device. The inset graph shows the EL spectrum. Reproduced with permission from ref. 36. Copyright 2014 John Wiley & Sons.

In 2019, Su et al. synthesized three orange-yellow to NIR hot exciton emitters, BTDF-TPA, BTDF-TTPA, and BTDF-TtTPA (Fig. 22).⁷⁷ By introducing thiophene as a π bridge into the D–A–D framework, the conjugation length of BTDF-TTPA and BTDF-TtTPA was extended, and the dihedral angle between the donor and acceptor was reduced, which led to a high FMO overlap. Therefore, the emission peak positions of the two compounds were red shifted 67 nm relative to BTDF-TPA without changing the PLQY. The maximum EQEs of BTDF-TTPA and BTDF-TtTPA were 5.75% and 4.94%, respectively. The non-doped device based on the latter achieved a maximum EQE of 1.44% peaking at 690 nm in NIR emission.

In 2019, Chi et al. reported two red emitters, TPABTPA and TPABCHO, based on the hot exciton mechanism.⁷⁸ According to theoretical calculations, the triplet excitons of TPABTPA and TPABCHO attenuated mainly through T₂/T₃ → S₁. TPABTPA and TPABCHO enabled non-doped OLEDs with excellent EQEs of 11.1% and 5.0% (Fig. 26), attributing to the high EUE of 82% and 46%, respectively. Such a high non-doped EQE was rarely observed in red OLEDs. They could also be utilized in white OLEDs. After appropriate device structure design, pure fluorescent WOLEDs were obtained with a maximum EQE of 23.0% and relatively stable white light emission CIE coordinates of (0.34, 0.33). As one of the best results, the NZ unit also possessed a large T₂–T₁ energy gap and pure fluorescent WOLEDs, and it provided a method to achieve HLCT-TADF pure fluorescent WOLEDs with high performance and similar orbital symmetry between T₁ and T₂, which can effectively inhibit IC from T₂ to T₁. Thus, it is the most widely used building block for deep-red and NIR hot exciton materials. In 2013, Ma's team reported a deep red emitter, TPA-NZP (Fig. 27), with intercrossed LE and CT characters, demonstrated by the solvatochromic experiment and density functional theory (DFT) calculations.⁷⁹ The PL properties and lifetime measurements of TPA-NZP in different solvents showed that the excited state of TPA-NZP in medium polar solvents was a HLCT state. Using TPA-NZP as the active layer in OLEDs, high EUE values of 93% (at low brightness) and 50% (at excitons) were achieved in the OLEDs. The emission peak of non-doped OLEDs was at 664 nm with a CIE of (0.67, 0.32), and the maximum EQE reached 2.8%. This work provided an important design idea for future HLCT materials with red or even deeper colors.

Fig. 26 EQE–luminance curves (inset: EL spectra) of non-doped TPABTPA and TPABCHO devices. Reproduced with permission from ref. 78. Copyright 2019 American Chemical Society.

Fig. 27 Molecular structures of the materials containing NZ units.

Yang and colleagues subsequently reported the hot exciton material TPA-NZC (Fig. 27).⁸⁰ Compared with TPA-NZP, TPA-NZC contained an additional cyanide group as a co-acceptor, which enhanced the acceptor capacity of NZ in D–A molecules. Compared with TPA-NZP (668 nm and PLQY of 15%), TPA-NZC exhibited an NIR emission of 710 nm with a large red-shift, as well as a maintained PLQY of 17% in films. Quantum chemical calculations and photophysical characterization showed that TPA-NZC exhibited HLCT characteristics, which contributed to the better performance of OLEDs. The non-doped OLEDs showed NIR emission peaking at 702 nm with a maximum EQE of 1.2%. This work demonstrated that the cyano group could be used as a co-acceptor in the preparation of D–A narrowband NIR materials. In 2017, Yang et al. reported an HLCT-AIE emitter, NZ2TPA (Fig. 27), introducing a DPA block as a donor on the basis of TPA-NZP to fabricate symmetric D–π–A–π–D materials.⁸¹ Attributed to its HLCT mechanism and AIE characteristics, NZ2TPA acquired an unprecedentedly high PLQY of 60% in the neat film, which was the highest among the reported organic small-molecule NIR emitters and even exceeded those of most phosphorescent NIR materials. The maximum EQE of 3.9% and an emission peak at 696 nm were achieved in the undoped device (Fig. 28), and the brightness reached 6330 cd m⁻². Thanks to the HLCT mechanism, at 1000 cd m⁻² brightness, it still showed an EQE of 2.8% with very low roll-off. This work provided a pathway for efficient D–π–A–π–D NIR OLEDs.

Fig. 28 (a) The structures of the NZ2TPA-based device as well as the energy levels of the used materials. (b) EQE versus brightness (inset: EL spectrum at 8 V). (c) Curve of EQE versus brightness (inset: EL spectrum at 7 V). (d) J–V–L characteristics. Reproduced with permission from ref. 81. Copyright 2017 John Wiley & Sons.

Subsequently, in 2017, they reported another HLCT-AIE material, NZ2AC (Fig. 27), also with a D–π–A–π–D structure by replacing the donor of NZ2TPA with DMAC.⁸² The NZ2AC-based doped device achieved a maximum EQE of 6.2% with an emission peak at 612 nm. Due to the AIE properties of NZ2AC, the non-doped OLEDs also achieved a deep red emission at 663 nm and a maximum EQE of 2.8%, corresponding to a maximum EUE of 93%. Because of the simultaneous HLCT and AIE features, both the doped and non-doped devices exhibited low efficiency roll-offs at high brightness, with EQEs of 3.0% and 2.3% at 5000 cd m⁻², respectively.

In 2017, Yang et al. published two NIR emitters, PXZ-3-NZP and PXZ-10-NZP (Fig. 27), which were composed of donor and acceptor moieties as well as 10-substituted isomeride PXZ-10-NZP for the purpose of comparison and understanding the essential difference of their excited state properties.⁸³ As a comparison, PXZ-10-NZP showed a far inferior performance to that of PXZ-3-NZP in both non-doped and doped devices, due to the instinct CT character of its S₁ excited state. The EL peak of the undoped device was located at 738 nm and the maximum EQE was 0.82%. The wavelength of this undoped device was one of the records of NIR fluorescent materials.

High-efficiency organic deep red/NIR emitters with desirable PLQYs and satisfactory EUE were still limited. In 2019, Tang and co-workers prepared three deep red/NIR D–A–D* emitters, TNZPPI, TNZtPPI and TNZ2tPPI, by using PPI/TPA as the asymmetric donor and NZ as the acceptor. The tert-butyl was introduced to regulate the molecular interactions.⁷³ Their interesting characterization of hybrid excited states containing tuned local excited and CT components was confirmed, and the effective high-lying RISC channel could be activated because of their large T₂–T₁ energy gap and small T₄–S₂ energy splitting, which matched well with the hot exciton mechanism. Among them, the doped device based on TNZPPI displayed the best performance with a deep red emission of 648 nm, a maximum EQE of 6.83% and an EUE of 82%. The electronic effect of the molecule structure and intermolecular interactions were all relative to their performance, which is very important for the design of high-efficiency NZ-based OLED materials.

2.4 PTZ based emitters

PTZ, as a classical electron-rich group, was first used in dye-sensitized solar cells.^84,85It has been widely applied as an electron donor for the construction of efficient organic light emitting materials.^36,58 A large number of PTZ-based derivatives show multiple molecular conformations, leading to different aggregation states and intermolecular interactions in the solid state. These features also entitle PTZ-based derivatives with some interesting characteristics such as AIE activity, piezochromic properties, etc.

In 2014, Adachi et al. developed a material, PTZ-TRZ (Fig. 29), containing the PTZ donor unit and 2,4,6-triphenyl-1,3,5-triazine (TRZ), the acceptor unit, which exhibited TADF properties.⁸⁶ Resulting from the distortion of PTZ, there existed two ground-state conformers with different energy gaps between the S₁ and T₁ states (1.14 and 0.18 eV). The PL spectra of PTZ-TRZ in toluene showed two broad, structureless emissions, which were attributed to intramolecular charge-transfer fluorescence from the quasi-axial conformation and quasi-equatorial conformation of PTZ (Fig. 30), respectively. The maximum EQE of the doped OLED based on PTZ-TRZ reached 10% with dual ICT fluorescence. The work reconfirmed the importance of structural chemical control in TADF materials. In 2018, Chou et al. reinvestigated the photophysical properties of PTZ-TRZ.⁸⁷ They explained that the complicated excitation behavior in toluene, in part, was due to the UV absorption cut-off region for toluene where the <315 nm excitation was greatly distorted by solvent absorption. Switching the solvent to cyclohexane with the UV cut-off wavelength at 235 nm simplified the results. In cyclohexane, quasi-axial and quasi-equatorial conformers existed for PTZ-TRZ in the ground state (Fig. 31). Upon electronic excitation, both conformers underwent structural relaxation to an energy minimum state where the PTZ was in a planar configuration.

Fig. 29 Molecular structures of the materials containing PTZ units.

Fig. 30 Schematic diagram of the one indirect excitation pathway of λE_PLs of PTZ-TRZ. Reproduced with permission from ref. 86. Copyright 2014 American Chemical Society.

Fig. 31 (a) Molecular structures. (b) HOMO and LUMO of the ground-state quasi-equatorial and quasi-axial conformers of PTZ-TRZ calculated at the CAM-B3LYP/cc-pVDZ level, respectively. (c) Schematic energy level diagram of PTZ-TRZ in cyclohexane solution exhibiting excited-state structural relaxation. Reproduced with permission from ref. 86. Copyright 2014 American Chemical Society and ref. 87. Copyright 2018 American Chemical Society.

In 2014, in order to solve the serious roll-off problem in the solid state, Chi et al. designed two TADF compounds SF2P and SFPC (Fig. 29), using DPS as the acceptor part, and PTZ and Cz as the donor, respectively.⁸⁸SFPC exhibited artful packing and highly ordered alignment without any π–π interactions, providing a high delayed fluorescence PLQY value of 93.3% in the solid state. The ΔE_ST values of SF2P and SFPC were estimated to be 0.03 eV and 0.20 eV, respectively. At the same time, SFPC could also show strong ML without treatment. Thus, exploiting a design and synthetic strategy to integrate the features of TADF, AIE, and ML into one compound was successfully achieved.

Chi et al. then explored another asymmetric D–A–D′ molecule similar to SFPC, OPC (Fig. 29), whose acceptor part was replaced by DPS with benzophenone.⁸⁹OPC was derived from two parent molecules, dicarbazolyl-substituted benzophenone (O2C)⁹⁰ and diphenothiazinyl-substituted benzophenone (O2P)⁹¹ with inherited emission from O2C and O2P (Fig. 29). Delayed fluorescence at 580 nm of OPC was confirmed according to temperature dependence measurements (Fig. 32). This finding provided a feasible molecular strategy to design AIDF white light-emitting organic molecules.

Fig. 32 PL spectra of SF2C and SF2P (a), SFPC (b), O2C and O2P (d), and OPC (e) in THF/water mixtures with different f_w. The HOMO, LUMO, and conformations of SF2P, SFPC, and SF2C (c), and O2C, O2P, and OPC (f) were optimized and calculated at B3LYP/6-31G(d,p). Reproduced with permission from ref. 89. Copyright 2015 John Wiley & Sons.

Chi et al. continually explored the dual-emissive compound, SCP (Fig. 29), according to the same design principle.⁹²SCP exhibited remarkable and linearly tunable mechanochromism and bright white-light emission with TADF character by fully inheriting the photophysical properties of the parent molecules SC2 and SP2. The PL spectrum showed the deep blue emission at 415 nm from SC2 and the yellow-green emission at 545 nm from SP2, respectively (Fig. 33). These observations demonstrated that creating asymmetric molecules following the principle of molecular heredity holds promise as a strategy for the development of functional materials.

Fig. 33 (a) PL spectra of SC2 and SP2 in the solid state. (b) Changes of PL spectra of SCP under grinding.

In 2017, on the basis of POZ-DBPHZ,⁹³ Takeda et al. developed U-shaped (D–A–D) conjugated multi-functional molecules, DBPHZ1 and DBPHZ2 (Fig. 29), comprising DBPHZ as an acceptor and PTZ as a donor.⁹⁴DBPHZ1 and DBPHZ2 showed a significant response to a variety of external stimuli because of the “two-conformation-switchable” PTZ units. Time-resolved photophysical measurements revealed that DBPHZ1 possessed a very small ΔE_ST induced by efficient HOMO/LUMO decoupling and exhibited strong TADF. Finally, the doped device based on DBPHZ1 achieved a maximum EQE of 16.8% with an electroluminescent peak at 613 nm.

For TADF-based OLEDs, the problem of serious efficiency roll-off at high brightness is an obstacle for their commercialization. The intermolecular electron-exchange interaction, as described by Dexter energy transfer (DET), dominates the concentration quenching of DF molecules.⁹⁵ Owing to the short-range nature of DET, high-concentration exciton annihilation can be effectively relieved in loosely aligned luminogenic molecules with weak intermolecular interactions. In recent years, the AIDF materials have shown the potential to provide a possible solution to reduce the efficiency roll-off. Based on this concept, Tang et al. designed and synthesized three compounds, CP-BP-PXZ, CP-BP-PTZ (Fig. 29) and CP-BP-DMAC.⁴¹ All of them showed obvious AIE characteristics, with PLQYs of 58%, 45.3% and 67.4% in neat films, respectively. The undoped CP-BP-PTZ device displayed a heavy efficiency roll-off which might be due to the unbalanced charge transport of CP-BP-PTZ as inferred from the current density–voltage curves of its pure hole- and electron-only devices. The non-doped OLEDs of these AIDF materials provided lower turn-on voltages and higher brightness than doped devices. Significantly high CE, PE and EQE up to 59.1 cd A⁻¹and 65.7 lm W⁻¹ and 18.4% were obtained in CP-BP-PXZ undoped devices, comparable to doped devices. The non-doped devices showed the obvious merit of smaller efficiency roll-off compared with the doped ones.

Subsequently, in 2018, Tang et al. published another molecule PTZ-AQ (Fig. 29) which exhibited polymorphism, AIE, MCL, and TADF characteristics.⁹⁶ By controlling the precipitation conditions, five different aggregates were obtained with emission from green to deep red (Fig. 34). Among them, the R-crystal exhibited considerable TADF properties, with a high PLQY of 84.8% and a small ΔE_ST of 0.01 eV. This work further provided the possibility of achieving multifunctional OLED materials.Li's group reported an ML PTZ derivative of FCO-CzS (Fig. 29), in which the D–A structure further enhanced the dynamic electronic properties and the fluorine atom increased the intermolecular interactions and restricted non-radiative transitions, promoting PL and ML emissions.⁹⁷FCO-CzS showed AIE activity and variable mechanical luminescence ranging from blue to white and yellow under continuous mechanical stimulation. The single crystal structure analysis and theoretical calculations as well as the comparison with FCO-CzO indicated that the source of dynamic mechanical luminescence was related to the multiple conformations of PTZ. This finding revealed the close relationship between the molecular conformation and ML process, which further deepened the understanding about the ML properties of PTZ derivatives.

Fig. 34 (a) The molecular structure of PTZ-AQ. (b) Photographs of the five solid states of PTZ-AQ (left: under ambient light; right: under UV excitation; scale bar: 200 mm). (c) PL spectra of PTZ-AQ in different aggregation states. Reproduced with permission from ref. 96. Copyright 2018 John Wiley & Sons.

PTZ exhibits dual stable conformations universally and showed the probability of exhibiting dual emissions during excitation, suggesting its great potential in the development of single emitters for highly efficient WOLEDs. In 2018, Li and colleagues proposed a series of single white emitters to simplify the device structure.⁹⁸ Four emitters PTZ-TTR, PTZ-Ph-TTR (Fig. 29), DMAC-TTR, and DMAC-Ph-TTR which could theoretically exhibit dual stable conformations and TADF characteristics were achieved. Two emission peaks were observed in their PL spectra deriving from quasi-axial conformation and quasi-equatorial conformation. Their band gaps and relative distributions could be well adjusted by changing the donor moieties with different flexibilities. The relative distribution of dual conformations could also be well tuned by controlling the interaction between the donor and acceptor segments. Through careful optimization, the PTZ-TTR based WOLEDs exhibited standard white light emission with CIE coordinates of (0.33, 0.33) and a high CRI value of 92. The device based on PTZ-Ph-TTR achieved an emission close to that of warm white light and a high maximum forward-viewing EQE of 16.34% (Fig. 35).

Fig. 35 (a) PE-luminance-EQE curves of PTZ-TTR and PTZ-Ph-TTR. (b) Luminance–voltage–current density curves and normalized EL spectra of PTZ-TTR and PTZ-Ph-TTR. (c) Normalized EL spectra of PTZ-TTR (inset): CIE variations from 100 to 1000 cd m⁻². (d) Normalized EL spectra of PTZ-Ph-TTR (inset: CIE variations from 100 to 10000 cd m⁻² and observed emissions under different driving voltages)-based WOLEDs. Reproduced with permission from ref. 98. Copyright 2018 American Chemical Society.

In 2020, Zhang et al. reported two single white emitters, 2PQ-PTZ and 4PQ-PTZ (Fig. 29), considering the distinct advantages of PTZ conformation isomers.⁹⁹ Two distinct emission bands from the quasi-axial and quasi-equatorial conformation isomers of PTZ were observed in both 2PQ-PTZ and 4PQ-PTZ in dilute solutions of 10^-4 M toluene. According to X-ray crystallography analysis, 2PQ-PTZ possessed a pair of quasi-axial and quasi-equatorial isomers with an accurate molar ratio of 1 : 1. A 2PQ-PTZ-based WOLED with CIE coordinates of (0.32, 0.34) and color rendering index (CRI) of 89 was demonstrated. The WOLEDs exhibited a high EQE of 10.12% at a lower doping ratio of 1 wt%. With the increase of doping ratio to 18%, a record-high EQE of 25.0% in PTZ-based TADF OLEDs was also obtained.

With the same design strategy, in 2019, Wang et al. reported an organic molecule PTZ-BP (Fig. 29) consisting of a PTZ donor and a BP acceptor.¹⁰⁰ The PL spectra of PTZ-BP in diluted toluene solution at room temperature and low temperature showed two maximum emission values at 440 nm and 550 nm, respectively, and the emission at 550 nm showed typical TADF characteristics. The X-ray structure of the single crystal and PL spectra showed that the two emission peaks were derived from the H-extra and H-intra conformations of PTZ-BP, respectively. The doped OLEDs based on the 5,5′-di(9H-carbazol-9-yl)-3,3′-bipyridine (DczDPy, Fig. 20): PTZ-BP system showed good properties with a maximum EQE of 6.2%. Furthermore, they also used CBP instead of DczDPy as a host to fabricate an EL device with the same device scheme. The device with the PTZ-BP:CBP-doped film (4 wt%) as the emitter layer showed a maximum CE of 15.5 cd A⁻¹ and an EQE of 5.44%, respectively.

As mentioned above, it is an effective solution to reduce the intermolecular interactions between molecules in order to achieve efficient TADF with reduced efficient roll-off. In 2019, Lu's group synthesized a TADF molecule, PTZMes₂B (Fig. 29), by using dimethylarylborane (Mes₂B) as a space-repelling electron acceptor and PTZ as a donor.¹⁰¹ XRD analysis showed that the donor and acceptor were connected in a torsional way with a dihedral angle of 87.83°. The PL spectrum of the undoped film showed a PLQY of 65% and ΔE_ST of 0.18 eV. The maximum EQE of the undoped device based on PTZMes₂B was 19.66%, and the EQE could remain at 17.31% at a high luminance of 1500 cd m⁻².

In 2019, Chen et al. reported three TADF emitters, DMAC-ND, PTZ-ND (Fig. 29) and PXZ-ND.¹⁰² The single crystal of PTZ-ND for X-ray diffraction analyses indicated that the torsion angles between the two PTZ units and the naphthalene part in PTZ-ND were 66.34° and 65.31°, respectively. The distorted structure inhibited the aggregation and quenching of the emitter, and the weak intermolecular interactions limited the intramolecular movement and maintained the molecular rigidity. The doped device based on PTZ-ND showed a green emission with a maximum CE of 40.9 cd A⁻¹, a PE of 31.4 lm W⁻¹ and an EQE of 13.0%.

In 2020, they further reported three AIDF compounds, pipd-BZ-PXZ, pipd-BZ-PTZ (Fig. 29) and pipd-BZ-DMAC, where a strong electron-withdrawing group imidazo[1,2-a]pyridin-2-yl(phenyl)methanone (pipd) served as the electron acceptor and PXZ, PTZ, and DMAC functioned as electron donors.¹⁰³ Generated by a strong intramolecular CT effect, pipd-BZ-PXZ and pipd-BZ-PTZ exhibited orange to red emission in solution and neat films. The 6 wt% doped device of pipd-BZ-PTZ showed the best results: a maximum CE of 55.41 cd A⁻¹, a maximum PE of 58.03 lm W⁻¹ and a maximum EQE of 15.77%. At a 20 wt% doping concentration, the device performance was reduced, but the efficiency roll-off was greatly improved. This work once again demonstrated the feasibility of designing AIDF materials in achieving a low efficiency roll-off.

3. Conclusion and outlook

In summary, we have reviewed a variety of efficient sulphur-containing materials classified based on different functional groups including DBT, DPS, BZ/NZ and PTZ units. Sulphur possesses two lone pairs of electrons and can form sulfoxide or sulfone functional groups with electron-deficient characteristics. Therefore, sulphur-based fluorophores are promising materials for high-efficiency OLEDs. Their basic electronic structure, spatial structure and photophysical properties are summarized. To achieve OLEDs with excellent performance, all singlet and triplet excitons should be utilized for radiation luminescence in the device. The two main mechanisms for obtaining 100% IQE in pure organic small molecules, TADF and HLCT, have also been illustrated. DBT with an electron-rich nature is usually used in TADF molecules as an assistant donor and is responsible for generating π–π interactions to improve the EL properties. DPS derivatives are widely applied in green-blue OLEDs, mainly as an acceptor fragment of TADF materials. Several devices based on this maintain the highest level of OLED performance. BZ and NZ serve as the classical acceptors for HLCT materials, and a large number of OLEDs with good device performance, simplified device structures and small efficiency roll-offs are obtained. However, compared with TADF molecules, OLEDs based on BZ and NZ derivatives still have much room for further improvement in device efficiency. PTZ with multiple conformations is commonly utilized both in hot exciton and TADF materials. They also show potential applications in single-molecule based WOLEDs. These discussed sulphur-containing materials all cross the theoretical limited EQE (5%) by harvesting triplet excitons to the singlet state through RISC. Some special phenomena found in S-containing materials such as AIE and ML properties are also demonstrated. We believe that our summarization can be used to guide the choice of new sulphur-containing efficient emitters and facilitate their applications in OLEDs. Further research and the development of new sulphur-containing derivatives is ongoing to achieve breaking of the theoretical limited EQE while maintaining stable CIE color coordinates and low efficiency roll-off to elongate the operation lifetime.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This research is supported by the National Natural Science Foundation of China (22075100, 91833304).

References

1. C. W. Tang and S. A. Van Slyke, Appl. Phys. Lett., 1987, 51, 913–915.
2. L. S. Hung and C. H. Chen, Mater. Sci. Eng., R, 2002, 39, 143–222.
3. S. Reineke, Nat. Mater., 2015, 14, 459–462.
4. M. A. Baldo, D. F. O’Brien, M. E. Thompson and S. R. Forrest, Phys. Rev. B: Condens. Matter Mater. Phys., 1999, 60, 14422–14428.
5. L. Xiao, S.-J. Su, Y. Agata, H. Lan and J. Kido, Adv. Mater., 2009, 21, 1271–1274.
6. K. H. Kim, C. K. Moon, J. H. Lee, S. Y. Kim and J. J. Kim, Adv. Mater., 2014, 26, 3844–3847.
7. Y. C. Zhu, L. Zhou, H. Y. Li, Q. L. Xu, M. Y. Teng, Y. X. Zheng, J. L. Zuo, H. J. Zhang and X. Z. You, Adv. Mater., 2011, 23, 4041–4046.
8. H. Shin, S. Lee, K. H. Kim, C. K. Moon, S. J. Yoo, J. H. Lee and J. J. Kim, Adv. Mater., 2014, 26, 4730–4734.
9. K. Udagawa, H. Sasabe, C. Cai and J. Kido, Adv. Mater., 2014, 26, 5062–5066.
10. K. H. Kim, S. Lee, C. K. Moon, S. Y. Kim, Y. S. Park, J. H. Lee, J. Woo Lee, J. Huh, Y. You and J. J. Kim, Nat. Commun., 2014, 5, 4769.
11. M. A. Baldo, D. F. O’Brien, Y. You, A. Shoustikov, S. Sibley, M. E. Thompson and S. R. Forrest, Nature, 1998, 395, 151–154.
12. D. Chen, W. Li, L. Gan, Z. Wang, M. Li and S.-J. Su, Mater. Sci. Eng., R, 2020, 142, 100581.
13. S. Sinha and A. P. Monkman, Appl. Phys. Lett., 2003, 82, 4651–4653.
14. S. Sinha, C. Rothe, R. Guntner, U. Scherf and A. P. Monkman, Phys. Rev. Lett., 2003, 90, 127402.
15. C.-J. Chiang, A. Kimyonok, M. K. Etherington, G. C. Griffiths, V. Jankus, F. Turksoy and A. P. Monkman, Adv. Funct. Mater., 2013, 23, 739–746.
16. W. Li, D. Liu, F. Shen, D. Ma, Z. Wang, T. Feng, Y. Xu, B. Yang and Y. Ma, Adv. Funct. Mater., 2012, 22, 2797–2803.
17. W. Li, Y. Pan, R. Xiao, Q. Peng, S. Zhang, D. Ma, F. Li, F. Shen, Y. Wang, B. Yang and Y. Ma, Adv. Funct. Mater., 2014, 24, 1609–1614.
18. C. Lin, P. Han, S. Xiao, F. Qu, J. Yao, X. Qiao, D. Yang, Y. Dai, Q. Sun, D. Hu, A. Qin, Y. Ma, B. Z. Tang and D. Ma, Adv. Funct. Mater., 2021, 2106912,  DOI:10.1002/adfm.202106912.
19. H. Kaji, H. Suzuki, T. Fukushima, K. Shizu, K. Suzuki, S. Kubo, T. Komino, H. Oiwa, F. Suzuki, A. Wakamiya, Y. Murata and C. Adachi, Nat. Commun., 2015, 6, 8476.
20. H. Tanaka, K. Shizu, H. Miyazaki and C. Adachi, Chem. Commun., 2012, 48, 11392–11394.
21. J. Jortner, S. I. Choi, J. L. Katz and S. A. Rice, Phys. Rev. Lett., 1963, 11, 323–326.
22. D. Y. Kondakov, Philos. Trans. R. Soc., A, 2015, 373, 20140321.
23. H. Uoyama, K. Goushi, K. Shizu, H. Nomura and C. Adachi, Nature, 2012, 492, 234–238.
24. X. Cai and S.-J. Su, Adv. Funct. Mater., 2018, 28, 1802558.
25. W.-C. Chen, C.-S. Lee and Q.-X. Tong, J. Mater. Chem. C, 2015, 3, 10957–10963.
26. R. A. Keller, Chem. Phys. Lett., 1969, 3, 27–29.
27. S. Reindl and A. Penzkofer, Chem. Phys., 1996, 211, 431–439.
28. R. W. Redmond, I. E. Kochevar, M. Krieg, G. Smith and W. G. Mcgimpsey, J. Phys. Chem. A, 1997, 101, 2773–2777.
29. H. Fukumura, K. Kikuchi, K. Koike and H. Kokubun, J. Photochem. Photobiol., A, 1988, 42, 283–291.
30. P. T. Chou, M. L. Martinez and S. L. Studer, J. Phys. Chem., 1991, 95, 10306–10310.
31. M. L. Martinez, S. L. Studer and P. T. Chou, J. Am. Chem. Soc., 1990, 112, 2427–2429.
32. X. Qiu, Y. Xu, C. Wang, M. Hanif, J. Zhou, C. Zeng, Y. Li, Q. Jiang, R. Zhao, D. Hu and Y. Ma, J. Mater. Chem. C, 2019, 7, 5461–5467.
33. J.-Y. Hu, Y.-J. Pu, F. Satoh, S. Kawata, H. Katagiri, H. Sasabe and J. Kido, Adv. Funct. Mater., 2014, 24, 2064–2071.
34. C. Li, M. Hanif, X. Li, S. Zhang, Z. Xie, L. Liu, B. Yang, S. Su and Y. Ma, J. Mater. Chem. C, 2016, 4, 7478–7484.
35. W. Z. Yuan, X. Bin, G. Chen, Z. He, J. Liu, H. Ma, Q. Peng, B. Wei, Y. Gong, Y. Lu, G. He and Y. Zhang, Adv. Opt. Mater., 2017, 5, 1700466.
36. L. Yao, S. Zhang, R. Wang, W. Li, F. Shen, B. Yang and Y. Ma, Angew. Chem., Int. Ed., 2014, 53, 2119–2123.
37. H. Mutlu, E. B. Ceper, X. Li, J. Yang, W. Dong, M. M. Ozmen and P. Theato, Macromol. Rapid Commun., 2019, 40, 1800650.
38. R. A. Keller, Chem. Phys. Lett., 1969, 3, 27–29.
39. S. Reindl and A. Penzkofer, Chem. Phys., 1996, 211, 431–439.
40. R. W. Redmond, I. E. Kochevar, M. Krieg, G. Smith and W. G. Mcgimpsey, J. Phys. Chem. A, 1997, 101, 2773–2777.
41. J. Huang, H. Nie, J. Zeng, Z. Zhuang, S. Gan, Y. Cai, J. Guo, S.-J. Su, Z. Zhao and B. Z. Tang, Angew. Chem., Int. Ed., 2017, 56, 12971–12976.
42. H. L. Shen, F. Huang, L. T. Hou, H. B. Wu, W. Cao, W. Yang and Y. Cao, Synth. Met., 2005, 152, 257–260.
43. J. A. Mikroyannidis, H. A. Moshopoulou, J. A. Anastasopoulos, M. M. Styllanakis, L. Fenenko and C. Adachi, J. Polym. Sci., Part A: Polym. Chem., 2006, 44, 6790–6800.
44. W. S. Huang, Y. H. Wu, Y. C. Hsu, H. C. Lin and J. T. Lin, Polymer, 2009, 50, 5945–5958.
45. J. Guo, X.-L. Li, H. Nie, W. Luo, S. Gan, S. Hu, R. Hu, A. Qin, Z. Zhao, S.-J. Su and B. Z. Tang, Adv. Funct. Mater., 2017, 27, 1606458.
46. J. Fan, L. Lin and C.-K. Wang, J. Mater. Chem. C, 2017, 5, 8390–8399.
47. J. Guo, X.-L. Li, H. Nie, W. Luo, R. Hu, A. Qin, Z. Zhao, S.-J. Su and B. Z. Tang, Chem. Mater., 2017, 29, 3623–3631.
48. R. Huang, N. A. Kukhta, J. S. Ward, A. Danos, A. S. Batsanov, M. R. Bryce and F. B. Dias, J. Mater. Chem. C, 2019, 7, 13224–13234.
49. R. Huang, J. S. Ward, N. A. Kukhta, J. Avó, J. Gibson, T. Penfold, J. C. Lima, A. S. Batsanov, M. N. Berberan-Santos, M. R. Bryce and F. B. Dias, J. Mater. Chem. C, 2018, 6, 9238–9247.
50. Q. Zhang, B. Li, S. Huang, H. Nomura, H. Tanaka and C. Adachi, Nat. Photonics, 2014, 8, 326–332.
51. X. Zeng, K.-C. Pan, W.-K. Lee, S. Gong, F. Ni, X. Xiao, W. Zeng, Y. Xiang, L. Zhan, Y. Zhang, C.-C. Wu and C. Yang, J. Mater. Chem. C, 2019, 7, 10851–10859.
52. M. Liu, R. Komatsu, X. Cai, H. Sasabe, T. Kamata, K. Nakao, K. Liu, S.-J. Su and J. Kido, Adv. Opt. Mater., 2017, 5, 1700334.
53. Q. Zhang, J. Li, K. Shizu, S. Huang, S. Hirata, H. Miyazaki and C. Adachi, J. Am. Chem. Soc., 2012, 134, 14706–14709.
54. S. Wu, M. Aonuma, Q. Zhang, S. Huang, T. Nakagawa, K. Kuwabara and C. Adachi, J. Mater. Chem. C, 2014, 2, 421–424.
55. Q. Zhang, D. Tsang, H. Kuwabara, Y. Hatae, B. Li, T. Takahashi, S. Y. Lee, T. Yasuda and C. Adachi, Adv. Mater., 2015, 27, 2096–2100.
56. L. Zhan, Z. Chen, S. Gong, Y. Xiang, F. Ni, X. Zeng, G. Xie and C. Yang, Angew. Chem., Int. Ed., 2019, 58, 17651–17655.
57. M. Liu, R. Komatsu, X. Cai, K. Hotta, S. Sato, K. Liu, D. Chen, Y. Kato, H. Sasabe, S. Ohisa, Y. Suzuri, D. Yokoyama, S.-J. Su and J. Kido, Chem. Mater., 2017, 29, 8630–8636.
58. Z. Yang, Z. Mao, C. Xu, X. Chen, J. Zhao, Z. Yang, Y. Zhang, W. Wu, S. Jiao, Y. Liu, M. P. Aldred and Z. Chi, Chem. Sci., 2019, 10, 8129–8134.
59. M. Liu, Y. Seino, D. Chen, S. Inomata, S. J. Su, H. Sasabe and J. Kido, Chem. Commun., 2015, 51, 16353–16356.
60. X. Tang, Q. Bai, Q. Peng, Y. Gao, J. Li, Y. Liu, L. Yao, P. Lu, B. Yang and Y. Ma, Chem. Mater., 2015, 27, 7050–7057.
61. C. Fu, S. Luo, Z. Li, X. Ai, Z. Pang, C. Li, K. Chen, L. Zhou, F. Li, Y. Huang and Z. Lu, Chem. Commun., 2019, 55, 6317–6320.
62. I. Lee and J. Y. Lee, Org. Electron., 2016, 29, 160–164.
63. Y. Luo, S. Li, Y. Zhao, C. Li, Z. Pang, Y. Huang, M. Yang, L. Zhou, X. Zheng, X. Pu and Z. Lu, Adv. Mater., 2020, 32, 2001248.
64. F. B. Dias, K. N. Bourdakos, V. Jankus, K. C. Moss, K. T. Kamtekar, V. Bhalla, J. Santos, M. R. Bryce and A. P. Monkman, Adv. Mater., 2013, 25, 3707–3714.
65. X. He, T. Shan, X. Tang, Y. Gao, J. Li, B. Yang and P. Lu, J. Mater. Chem. C, 2016, 4, 10205–10208.
66. H. Wang, L. Xie, Q. Peng, L. Meng, Y. Wang, Y. Yi and P. Wang, Adv. Mater., 2014, 26, 5198–5204.
67. L. Xie, G. Han, Y. Chen, H. Wang, X. Kong, X. Wei, J. Liu, Y. Yi, B. Chen, P. Wang and Y. Wang, J. Mater. Chem. C, 2016, 4, 10776–10780.
68. X. Wei, Y. Chen, R. Duan, J. Liu, R. Wang, Y. Liu, Z. Li, Y. Yi, Y. Yamada-Takamura, P. Wang and Y. Wang, J. Mater. Chem. C, 2017, 5, 12077–12084.
69. X. Wei, Z. Li, T. Hu, R. Duan, J. Liu, R. Wang, Y. Liu, X. Hu, Y. Yi, P. Wang and Y. Wang, Adv. Opt. Mater., 2019, 7, 1801767.
70. S. P. Xiang, Z. Huang, S. Q. Sun, X. L. Lv, L. W. Fan, S. F. Ye, H. T. Chen, R. D. Guo and L. Wang, J. Mater. Chem. C, 2018, 6, 11436–11443.
71. J. Rao, C. Zhao, Y. Wang, K. Bai, S. Wang, J. Ding and L. Wang, ACS Omega, 2019, 4, 1861–1867.
72. Y. P. Jeon, B. K. Kong, E. J. Lee, K.-H. Yoo and T. W. Kim, Nano Energy, 2019, 59, 560–568.
73. Q. Wan, J. Tong, B. Zhang, Y. Li, Z. Wang and B. Z. Tang, Adv. Opt. Mater., 2019, 8, 1901520.
74. W. Li, Y. Pan, L. Yao, H. Liu, S. Zhang, C. Wang, F. Shen, P. Lu, B. Yang and Y. Ma, Adv. Opt. Mater., 2014, 2, 892–901.
75. C. Wang, X. Li, Y. Pan, S. Zhang, L. Yao, Q. Bai, W. Li, P. Lu, B. Yang, S. Su and Y. Ma, ACS Appl. Mater. Interfaces, 2016, 8, 3041–3049.
76. J. Liu, Z. Li, T. Hu, X. Wei, R. Wang, X. Hu, Y. Liu, Y. Yi, Y. Yamada-Takamura, Y. Wang and P. Wang, Adv. Opt. Mater., 2019, 7, 1801190.
77. W. Xie, B. Li, X. Cai, M. Li, Z. Qiao, X. Tang, K. Liu, C. Gu, Y. Ma and S. J. Su, Front. Chem., 2019, 7, 276.
78. X. Chen, Z. Yang, W. Li, Z. Mao, J. Zhao, Y. Zhang, Y. C. Wu, S. Jiao, Y. Liu and Z. Chi, ACS Appl. Mater. Interfaces, 2019, 11, 39026–39034.
79. W. Li, Y. Pan, R. Xiao, Q. Peng, S. Zhang, D. Ma, F. Li, F. Shen, Y. Wang, B. Yang and Y. Ma, Adv. Funct. Mater., 2014, 24, 1609–1614.
80. X. Tang, X.-L. Li, H. Liu, Y. Gao, Y. Shen, S. Zhang, P. Lu, B. Yang, S.-J. Su and Y. Ma, Dyes Pigm., 2018, 149, 430–436.
81. T. Liu, L. Zhu, C. Zhong, G. Xie, S. Gong, J. Fang, D. Ma and C. Yang, Adv. Funct. Mater., 2017, 27, 1606384.
82. T. Liu, L. Zhu, S. Gong, C. Zhong, G. Xie, E. Mao, J. Fang, D. Ma and C. Yang, Adv. Opt. Mater., 2017, 5, 1700145.
83. C. Wang, X.-L. Li, Y. Gao, L. Wang, S. Zhang, L. Zhao, P. Lu, B. Yang, S.-J. Su and Y. Ma, Adv. Opt. Mater., 2017, 5, 1700441.
84. H. N. Tian, X. C. Yang, R. K. Chen, Y. Z. Pan, L. Li, A. Hagfeldt and L. C. Sun, Chem. Commun., 2007, 3741–3743,  10.1039/b707485a.
85. Z. Xie, A. Midya, K. P. Loh, S. Adams, D. J. Blackwood, J. Wang, X. Zhang and Z. Chen, Prog. Photovoltaics, 2010, 18, 573–581.
86. H. Tanaka, K. Shizu, H. Nakanotani and C. Adachi, J. Phys. Chem. C, 2014, 118, 15985–15994.
87. D.-G. Chen, T.-C. Lin, Y.-A. Chen, Y.-H. Chen, T.-C. Lin, Y.-T. Chen and P.-T. Chou, J. Phys. Chem. C, 2018, 122, 12215–12221.
88. S. Xu, T. Liu, Y. Mu, Y. F. Wang, Z. Chi, C. C. Lo, S. Liu, Y. Zhang, A. Lien and J. Xu, Angew. Chem., Int. Ed., 2015, 54, 874–878.
89. Z. Xie, C. Chen, S. Xu, J. Li, Y. Zhang, S. Liu, J. Xu and Z. Chi, Angew. Chem., Int. Ed., 2015, 54, 7181–7184.
90. S. Y. Lee, T. Yasuda, Y. S. Yang, Q. Zhang and C. Adachi, Angew. Chem., Int. Ed., 2014, 53, 6402–6406.
91. C. J. Chen, J. Y. Liao, Z. G. Chi, B. J. Xu, X. Q. Zhang, D. B. Kuang, Y. Zhang, S. W. Liu and J. R. Xu, J. Mater. Chem., 2012, 22, 8994–9005.
92. B. Xu, Y. Mu, Z. Mao, Z. Xie, H. Wu, Y. Zhang, C. Jin, Z. Chi, S. Liu, J. Xu, Y.-C. Wu, P.-Y. Lu, A. Lien and M. R. Bryce, Chem. Sci., 2016, 7, 2201–2206.
93. P. Data, P. Pander, M. Okazaki, Y. Takeda, S. Minakata and A. P. Monkman, Angew. Chem., Int. Ed., 2016, 55, 5739–5744.
94. M. Okazaki, Y. Takeda, P. Data, P. Pander, H. Higginbotham, A. P. Monkman and S. Minakata, Chem. Sci., 2017, 8, 2677–2686.
95. J. Lee, N. Aizawa, M. Numata, C. Adachi and T. Yasuda, Adv. Mater., 2017, 29, 1604856.
96. B. Huang, W. C. Chen, Z. Li, J. Zhang, W. Zhao, Y. Feng, B. Z. Tang and C. S. Lee, Angew. Chem., Int. Ed., 2018, 57, 12473–12477.
97. J. Yang, J. Qin, P. Geng, J. Wang, M. Fang and Z. Li, Angew. Chem., Int. Ed., 2018, 57, 14174–14178.
98. K. Wang, Y.-Z. Shi, C.-J. Zheng, W. Liu, K. Liang, X. Li, M. Zhang, H. Lin, S.-L. Tao, C.-S. Lee, X.-M. Ou and X.-H. Zhang, ACS Appl. Mater. Interfaces, 2018, 10, 31515–31525.
99. B. Li, Z. Li, F. Guo, J. Song, X. Jiang, Y. Wang, S. Gao, J. Wang, X. Pang and L. Zhao, ACS Appl. Mater. Interfaces, 2020, 12, 14233–14243.
100. C. Li, J. Liang, B. Liang, Z. Li, Z. Cheng, G. Yang and Y. Wang, Adv. Opt. Mater., 2019, 7, 1801667.
101. X. Tang, Y. Tao, H. Liu, F. Liu, X. He, Q. Peng, J. Li and P. Lu, Front. Chem., 2019, 7, 373.
102. C. Chen, H.-Y. Lu, Y.-F. Wang, M. Li, Y.-F. Shen and C.-F. Chen, J. Mater. Chem. C, 2019, 7, 4673–4680.
103. Z. Yang, Y. Zhan, Z. Qiu, J. Zeng, J. Guo, S. Hu, Z. Zhao, X. Li, S. Ji, Y. Huo and S.-J. Su, ACS Appl. Mater. Interfaces, 2020, 12, 29528–29539.

---