Zhaohan
Li†
*ab,
Jiaojiao
Song†
b,
Anming
Li
a,
Huaibin
Shen
*b and
Zuliang
Du
b
aCollege of Physics and Electrical Engineering, Zhengzhou Normal University, Zhengzhou 450044, China. E-mail: lizhaohan22@163.com
bKey Laboratory for Special Functional Materials of Ministry of Education, National & Local Joint Engineering Research Center for High-Efficiency Display and Lighting Technology, Henan University, Kaifeng 475004, China. E-mail: shenhuaibin@henu.edu.cn
First published on 23rd January 2023
As the emitters of quantum dots (QDs) light-emitting diodes (QLEDs), QDs, which are responsible for the charge injection, charge transportation, and especially exciton recombination, play a significant role in QLEDs. With the crucial advances made in QDs, such as the advancement of synthetic methods and the understanding of luminescence mechanisms, QLEDs also demonstrate a dramatic improvement. Until now, efficiencies of 30.9%, 28.7% and 21.9% have been achieved in red, green and blue devices, respectively. However, in QLEDs, some issues are still to be solved, such as the imbalance of charge injection and exciton quenching processes (defect-assisted recombination, Auger recombination, energy transfer and exciton dissociation under a high electric field). In this review, we will provide an overview of recent advances in the study and understanding of the working mechanism of QLEDs and the exciton quenching mechanism of QDs in devices. Particular emphasis is placed on improving charge injection and suppressing exciton quenching. An in-depth understanding of this progress may help develop guidelines to direct QLED development.
As the emission centers of QLEDs, QDs play a crucial role in the efficiency, luminance and lifetime of devices. Also, with the important advances made in QDs, such as the deep understanding of ligand engineering and structure engineering,1,3,17 the performances of QLEDs have also been dramatically improved. Therefore, insights from these progresses will be helpful to develop a set of guidelines to direct QLED innovation. In this review, we will provide an overview on recent advances in the understanding of the working mechanism of QLED devices and the approaches to improve device performances.
Colors | Year | Device structure | EL (nm) | EQEmax (%) | L max (cd m−2) | V on (V) | Lifetime | Ref. |
---|---|---|---|---|---|---|---|---|
Red | 2014 | ITO/PEDOT:PSS/poly-TPD/PVK/QDs/PMMA/ZnO/Ag | 640 | 20.5 | 42000 | 1.7 | L 0 = 100 cd m−2, T50 > 100000 h | 7 |
2015 | ITO/PEDOT:PSS/TFB/QDs/ZnO/Al | 625 | 12.0 | 21000 | 1.5 | L 0 = 100 cd m−2, T50 > 300000 h | 18 | |
2018 | ITO/PEDOT:PSS/TFB/QDs/ZnO/Al | 631 | 15.1 | — | 1.7 | L 0 = 100 cd m−2, T50 > 2200000 h, L0 = 1000 cd m−2, T95 > 2300 h | 19 | |
2018 | ITO/PEDOT:PSS/poly-TPD/PVK/QDs/ZnMgO/Ag | 624 | 18.2 | — | 1.7 | L 0 = 100 cd m−2, T50 = 190000 h | 20 | |
2018 | ITO/ZnO + PVP/QDs/TCTA/MoOx/Al | ∼611 | 13.5 | — | 1.9 | L 0 = 100 cd m−2, T50 = 1330000 h, L0 = 1000 cd m−2, T50 = 23660 h | 21 | |
2019 | ITO/NiOx–BA–CF3/poly-TPD/QDs/ZnMgO/Ag | 625 | 13.4 | — | 1.65 | L 0 = 1000 cd m−2, T95 = 2500 h | 22 | |
2019 | ITO/PEDOT:PSS/TFB/QDs/ZnO/Al | 601 | 21.6 | 357000 | ∼1.92 | L 0 = 100 cd m−2, T50 > 1600000 h, L0 = 7000 cd m−2, T50 > 840 h | 23 | |
2019 | ITO/PEDOT:PSS/TFB/QDs/ZnO/Al | 602 | 30.9 | 334000 | 1.9 | L 0 = 100 cd m−2, T50 > 1800000 h, L0 = 2000 cd m−2, T50 > 7300 h | 15 | |
2020 | ITO/PEDOT:PSS/TFB/QDs/ZLMO@MO/Al | 636 | 20.6 | — | 1.7 | L 0 = 1000 cd m−2, T95 > 11000 h | 24 | |
2020 | ITO/PEDOT:PSS/poly-TPD/PVK/QDs/ZnMgO/Ag | — | 20.2 | — | 1.65 | L 0 = 1000 cd m−2, T95 = 3800 h | 25 | |
2020 | ITO/ZnO/QDs/CBP:BCBP/MoOx/C60/Al | ∼610 | 18.3 | 410000 | ∼2.56 | L 0 = 100 cd m−2, T70 = 2140000 h | 26 | |
2021 | Ag/ZnO/QD/CBP/MoOx/HAT–CN–Ag | 615 | 14.7 | 650000 | — | L 0 = 100 cd m−2, T50 = 12600000 h | 27 | |
2022 | ITO/PEDOT:PSS/TFB/QDs/ZnO/Al | ∼624 | 21.9 | — | 1.7 | L 0 = 1000 cd m−2, T95 > 21000 h | 28 | |
Green | 2015 | ITO/PEDOT:PSS/TFB/QDs/ZnO/Al | 537 | 14.5 | — | 2.0 | L 0 = 100 cd m−2, T50 > 90000 h | 18 |
2017 | ITO/PEDOT:PSS/TFB/QDs/ZnO/Al | 532 | 16.5 | 78000 | 2.2 | L 0 = 100 cd m−2, T50 > 480000 h | 29 | |
2018 | ITO/ZnO/PVK/QD/PEIE/poly-TPD/MoOx/Al | 525 | 22.4 | 72814 | 5.75 | — | 30 | |
2019 | ITO/PEDOT:PSS/TFB/QDs/ZnO/Al | ∼534 | 22.9 | 614000 | — | L 0 = 100 cd m−2, T50 > 1760000 h, L0 = 7000 cd m−2, T50 > 770 h | 23 | |
2020 | ITO/PEDOT:PSS/TFB/QDs/ZnO/Al | 530 | 23.9 | ∼13200 | ∼2.2 | L 0 = 100 cd m−2, T50 > 1655000 h, L0 = 1000 cd m−2, T95 > 2500 h | 31 | |
2022 | ITO/PEDOT:PSS/PF8Cz/QDs/ZnMgO/Al | 537 | 28.7 | — | 2.05 | L 0 = 100 cd m−2, T95 ∼ 580000 h, T50 ∼ 2570000 h | 16 | |
Blue | 2015 | ITO/PEDOT:PSS/PVK/QDs/ZnO/Al | 455 | 10.7 | 4000 | 2.6 | — | 18 |
2017 | ITO/PEDOT:PSS/PVK/QDs/ZnO/Al | 468 | 19.8 | 4890 | 5.1 | L 0 = 100 cd m−2, T50 = 47.4 h | 32 | |
2019 | ITO/PEDOT:PSS/TFB/QDs/ZnO/Al | ∼481 | 8.05 | 62600 | — | L 0 = 100 cd m−2, T50 > 7000 h, L0 = 7000 cd m−2, T50 > 6 h | 23 | |
2020 | ITO/PEDOT:PSS/TFB/QDs/ZnMgO/Al | 460 | 20.2 | 88900 | — | L 0 = 100 cd m−2, T50 = 15850 h | 33 | |
2022 | ITO/PEDOT:PSS/PF8Cz/QDs/ZnMgO/Al | 479 | 21.9 | — | 2.45 | L 0 = 100 cd m−2, T95 ∼ 4400 h, T50 ∼ 24000 h | 34 |
At present, the organic–inorganic hybrid structure is the most commonly used device structure. Also, most of the high-performance devices are based on the conventional organic–inorganic hybrid structure.15,18,23,29,35 Generally, the QLEDs with the conventional structure have 4 functional layers, as shown in Fig. 1, that is, a hole injection layer (HIL), hole transport layer (HTL), QD emitting layer (EML) and electron transport layer (ETL). Under the driving of an external electric field, the holes are injected into the HIL from the anode of QLEDs. Then, the holes go through the HTL and are injected into the EML. Similarly, the electrons go through the ETL and are injected into the EML. If the electrons and holes in the QDs recombine radiatively, the QLEDs will give out light.
Fig. 1 (a) Schematic illustration of QLEDs with the traditional structure. (b) Energy band diagram of QLEDs and the schematic illustration of charge injection and charge recombination in QLEDs. |
For QLEDs, EQE, which is an important performance parameter, is equal to the ratio between the number of photons emitted from the device and the number of carriers injected into the device.36,37 Generally, EQE depends on three factors: the fraction of carriers effectively injected into QDs, the fraction of excitons that deexcite radiatively, and the fraction of photons that effectively eject from devices. And it can be expressed by the following equation:38
ηEQE = γ × ηr × ηout |
It is known that most organic ligands act as bulky insulating barriers between QDs, hindering charge transport.40 The poor conductivity of emitting layers, which is highly related to the long-chain organic ligand, constrains the luminance and efficiency of QLEDs. In 2015, Shen et al. reported that the electron mobility and hole mobility of the QD film all increased by replacing the longer oleic acid (OA) ligand with shorter 1-octanethiol (OT) ligand, as shown in Fig. 2(a). More importantly, the ligand exchange promoted the charge balance of QLEDs, and the blue devices showed an unprecedented high EQE of 12.2%, as shown in Fig. 2(b).41 By exchanging the intrinsic ligand OA with a short chain ligand tris(mercaptomethyl)nonane (TMMN), the QLEDs based on QDs capped with TMMN showed much higher efficiency and luminance and much lower turn-on voltage than the devices based on QDs coordinated by OA.29 Small inorganic ligands are also advantageous for charge injection/transport in QD based devices. In 2018, the Sargent group implemented conductive halides in Zn chalcogenide-shelled QDs to improve carrier mobility. The resulting devices demonstrated a reduced turn-on voltage of 2.5 V and maximum luminance of 460000 cd m−2, which was the highest value reported thus far.42 In 2020, Kim et al. exchanged the native OA ligand of ZnTeSe/ZnSe/ZnS core/shell/shell (C/S/S) QDs with ZnCl2 through two steps of ligand exchange: a liquid-phase treatment (referred to as C/S/S–Cl(l)) and a film-washing treatment (referred to as C/S/S–Cl(f)). And they fabricated a QLED with a double EML consisting of C/S/S–Cl(l) and C/S/S–Cl(f) layers to improve charge injection/transport and recombination simultaneously. The resulting device showed an EQE of 20.2% and T50 lifetime of 15850 h at 100 cd m−2, which were the highest values reported thus far for blue QLEDs.33
Fig. 2 (a) Current density–voltage (J–V) characteristics of electron- and hole-only devices based on OA capped QDs and OT capped QDs; (b) EQE and current efficiency of the devices based on QDs with OA and OT ligands as a function of luminance (L). Reproduced with permission.41 Copyright 2015, American Chemical Society. (c) Ultraviolet photoelectron spectroscopy data of Zn1−xCdxSe/ZnSe/ZnS core/shell QDs with OA or EHT surface ligands; (d) current density and luminance versus bias for devices based on Zn1−xCdxSe/ZnSe/ZnS core/shell QDs with OA or EHT surface ligands. Reproduced with permission.15 Copyright 2019, WILEY-VCH. |
Surface ligands not only influence the conductivity between QDs, but also the electronic properties. Upon ligand coordination, generally, the energy level of QDs will shift to a higher or lower energy direction. The shifts of energy levels originate from the induced dipole at the ligand/QD interface and the intrinsic dipole of the ligand.43,44 Ligand-induced energy level shifts are proved to be an important means to control the electronic properties of QDs and to optimize the performance of QD based optoelectronic devices. By engineering the band alignment of the QDs through ligand treatments, the QD solar cell showed a certified efficiency of 8.55% in 2014.45 Much work in the QLED field has also demonstrated that the ligand treatments are of great benefit to energy level tuning and charge balance. By exchanging the native OA ligand with tris(mercaptomethyl)nonane (TMMN), the VBM of TMMN-capped QDs shifts to the higher energy direction; moreover, the charge injection and charge balance are greatly improved for the corresponding QLED devices.29 Similarly, by exchanging the ligand OA with 2-ethylhexane-1-thiol (EHT), as shown in Fig. 2(c), the VBM of EHT-capped QDs is 0.27 eV higher than that of OA-capped QDs. Consequently, the QLEDs based on EHT-capped QDs showed much better charge injection and higher EQE than the devices based on QDs capped with native OA ligand, as shown in Fig. 2(d).15
Fig. 3 Electronic energy levels of selected III–V and II–VI semiconductors using the valence-band offsets. Reproduced with permission.46 Copyright 2009, Wiley-VCH. |
Recently, many advances were made in QLEDs by employing ZnSe based QDs. In 2015, the Qian group18 reported a full series of blue, green and red QLEDs with the efficiencies all over 10% by elaborately tailoring the nanostructure of QDs. They synthesized two kinds of QDs with ZnSe-rich and CdS-rich intermediate shells, respectively. As the VBM of ZnSe is 0.2 eV higher than that of CdS, the hole injection barrier is reduced with a ZnSe-rich intermediate shell, leading to a much higher injection current density (V > 3 V) than the CdS-rich QD based device. Consequently, the green devices based on ZnSe-rich QDs exhibited a much higher EQE than CdS-rich QD based devices. Following this energy design strategy, in 2018 the Qian group19 further synthesized CdSe/Cd1−xZnxSe/ZnSe QDs (ZnSe-QDs) with a high VBM ZnSe shell, which favors efficient hole injection. In terms of the higher VBM of ZnSe-QDs, the valence band offset at the TFB/ZnSe-QD interface is much smaller than that at the TFB/ZnS-QD interface. This excellent alignment of the VBM and HOMO between ZnSe-QDs and TFB interface is helpful for significantly declining the hole injection barrier in the corresponding QLEDs. Finally, the devices based on these tailored ZnSe-QDs exhibited much extended operation lifetime (T50 > 2200000 h@100 cd m−2). The device lifetime is the key performance parameter for the commercialization of QLEDs. The device lifetime of QLEDs is usually characterized by T50 (or T95), defined as the time for the luminance to decrease to half (or 95%) of its initial luminance while operating at a constant current density. The instability of devices is mainly induced by the imbalanced charge injection. Chang et al.47 pointed out that the leaking of excessive electrons into the HTL would lead to irreversible degradation of devices and reduces device lifetime. In 2019, Chen et al.48 demonstrated that differing from red QLEDs, the poor lifetime of blue QLEDs originates from the fast degradation at the QD–ETL junction. Therefore, improvement of charge injection and the balance of charge injection are also favourable for device stability.
In 2019, our group also employed QDs with high VBM to fabricate QLEDs. And these red, green and blue CdSe/ZnSe QD based QLEDs demonstrate simultaneously high brightness and EQE (21.6%@13300 cd m−2, 22.9%@52500 cd m−2, and 8.05%@10100 cd m−2 for red, green and blue devices, respectively). The high performance of devices can mainly be attributed to the Se throughout the whole of QDs, which could reduce the hole injection barrier, enhance the charge balance effectively, and consequently improve the device performance.23
The devices with high performance are the result of the coordination of all the functional layers. Therefore, besides tailoring QDs to improve charge injection, optimization of other functional layers, such as tuning the CTL, cathode and adding an additional blocking layer,7,49–53 could also enhance the device performances. In 2014, the Peng group7 improved the balance of charge injection by inserting a PMMA insulating layer, and the resulting devices demonstrated a record-high efficiency of 20.5%. In 2019, the Tan group53 demonstrated that the charge injection could be balanced through the CTL doping strategy and the device lifetime was improved ∼3.5 times.
Relatively, designing new charge transporting materials or adding additional layers significantly increases the technical difficulty and cost of commercial production. Therefore, it is believed that QD structure design is the most direct, effective, convenient and low-cost method to improve device performance, and also is the most promising method to improve device performance.
Until now, great efforts have been adopted to reduce the defects of QDs, such as surface engineering and interfacial engineering. In 2016, the Peng group55 employed two ways—shell isolation and surface treatment—to battle the surface traps. They demonstrated that the electron traps of CdSe/CdS core/shell QDs could be readily isolated from the electron wavefunction of the excitons with more than ∼2 monolayers of the CdS shell. In general, the shell with a wide bandgap could isolate the surface traps from the wavefunction of excitons. According to this design strategy, the Peng group9 synthesized CdxZn1−xSe/ZnSe/ZnS core/shell QDs through two complementary steps by localizing exciton wavefunctions away from the inorganic–organic interface of QDs. They first synthesized uniform-alloy CdxZn1−xSe QDs with their physical size greater than the exciton diameter to confine the excitons away from the inorganic–organic interface. Subsequently, by epitaxially growing wide bandgap and high-quality shells, the interface effects on the excitons can be reduced to a negligible level. The as-synthesized QDs exhibited a record-low PL full width at half-maximum (16.3 for ensemble PL and 9.7 nm for single-dot PL).
Due to the lattice mismatch of the core and shell in QDs, the core/shell QDs usually endure many interfacial defects. As interfacial defect states are more accessible than surface defects for the excitons, generally, the interfacial defects play a much more important role in the photoelectric properties of QDs and devices. Due to the gradual change of lattice constants, the alloyed QDs will have relatively fewer interfacial defects. By applying complementary analytical techniques of electron microscopy and atom probe tomography, Chae et al. elucidated the internal structure and related atomic distribution of core/shell structured CdSSe/ZnS QDs in three dimensions, particularly at heterostructure interfaces. The CdxZn1−xS gradient inner shell between the CdSe core and ZnS outermost shell alleviates the lattice misfit strain at the interfaces, thereby enhancing PL QY and photostability to a greater extent than those of other single-shell structures.56
In one aspect, the defects in QDs will decrease the quantity of free carriers; in the other aspect, the trapped carriers will lead to the charging of QDs and the enhancement of nonradiative Auger recombination, which will be further discussed in section 5.1.
As the QDs in QLEDs are in the form of densely packed films, the performances of QLEDs are to a large extent restricted by the energy transfer processes. In 2012, Pal et al.62 synthesized CdSe/CdS core/shell QDs with different shell thicknesses to study the influence of shell thickness on device performance. They assessed the effect of increasing shell thickness on the energy transfer process through time-resolved PL spectra of QD solution and QD solid films. As shown in Fig. 5, with the increase of shell thickness, the PL lifetime of CdSe/CdS QDs increases monotonically. Moreover, the PL dynamics of thin-shell CdSe/CdS QDs measured at shorter excitation wavelength becomes significantly faster when passing from diluted solution to solid film. This is a signature of energy transfer process from QDs with a wider bandgap to QDs with a narrower bandgap. With the increase of shell thickness, the discrepancy between QD solution and QD film PL dynamics is progressively reduced. As shown in Fig. 5(d), for the CdSe/16CdS core/shell QD film, the spectral diffusion is completely suppressed. It is because the thick shell could act as a spacer between the interacting excitons in the neighboring CdSe/CdS core/shell QDs, suppressing distance-dependent interparticle interactions.
Fig. 5 PL decay curves of CdSe/CdS QD solution (gray lines) and film (red, black and green lines) with a (a) 4, (b) 8, (c) 13, and (d) 16 monolayer CdS shell. Inset: The PL spectrums of CdSe/CdS QD solution (solid black curves) and film (dashed gray curves). The PL decay curves are collected at the emission energies indicated by the arrows in the insets. Reproduced with permission.62 Copyright 2012, American Chemical Society. |
Many groups have reported that the thick shell of QDs could effectively suppress the energy transfer process. The experiment results demonstrated that the PL spectra of the ZnCdSe/ZnS QD solid film shift to the longer wavelength side in comparison with that of the QD solution. Moreover, with increasing the shell thickness, the extent of redshift decreases gradually. This also indicates that the thick shell could suppress the energy transfer process of the close-packed film.63 Similarly, Yang et al. reported that by increasing shell thickness, the ZnSe/ZnS/ZnS QD-based devices exhibited suppressed Förster resonance energy transfer (FRET) compared with ZnSe/ZnS QD-based devices.34
Meanwhile, the field-induced charge delocalization will lead to the enhancement of energy transfer, due to the decrease of spacing between the interacting excitons. Therefore, as shown in Fig. 6(a), the field-induced energy transfer can be suppressed by increasing the shell thickness or reducing the electric field across QDs.65 Moreover, at high enough electric fields, the coulombic binding of excitons can be overcome, resulting in dissociation and formation of free carriers.66,67 Xie et al. applied a reverse-biased QLED to study the electric field effect on QD PL. And they pointed out that the 99.5% reduction in PL was accomplished by a synergistic interplay of the quantum-confined Stark effect and field-induced exciton dissociation.66 The electric field across QDs is determined by the external electric field and the built-in field, which is induced by the accumulated charges at the interface of the QD layer and charge transport layer.65,68 The mismatched Fermi levels induced a diffusion potential (Vdiff), which will drive the electron transfer from QDs into the EML. As shown in Fig. 6(b), a pseudo-electric-field Ep = Vdiff/d is used to quantify the driving force of the electron diffusion. When the QD film and ZnMgO layer are in contact, as shown in Fig. 6(c), the electrons will accumulate at the ZnMgO side and holes at the QD side to make the Fermi levels aligned with each other. The accumulated charges can induce a built-in field (Eq), which will cancel out Ep. Therefore, the built-in field can partially cancel out the small applied bias. By reducing the effective electric field (at 2 V bias), a record-high EQE of 22.56% and luminance of 136090 cd m−2 were achieved in InP-based QLEDs by Li et al.65
Fig. 6 (a) Schematic illustration of the field-enhanced electron delocalization and the two feasible ways to alleviate its impacts on charge transfer. The energy level alignments of the InP/ZnSe layer and ZnMgO (b) when they are separated, and (c) when they are in close contact. Reproduced with permission.65 Copyright 2022, Wiley-VCH. |
From the results of recent advances in QLEDs, it is believed that the following aspects will be critical to enhance the device performance.
Footnote |
† These authors contributed equally to this work. |
This journal is © The Royal Society of Chemistry 2023 |