Colloidal quantum dots (CQDs) are highly valued by researchers due to their small size, typically ranging from 1 to 10 nm, and their exceptional optoelectronic properties. These properties include high photoluminescence quantum yield (PLQY), narrow emission peaks, high color purity, stability, and effective solution processing (Fig. 1(c))^{[1, 2]}. CQDs are particularly promising for use in light-emitting diodes (LEDs), lasers, sensors, and solar cells due to these advantages^[3−7].

Quantum dot light-emitting diodes (QLEDs) are devices that use semiconductor quantum dots (QDs) as their core material, based on the structure and technology of organic light-emitting diodes (OLEDs)^[8]. Since QLEDs technology was introduced in 1994, it has undergone significant advancements over more than two decades. These improvements are largely due to the development of high-quality QDs, innovative device structures, and a deeper understanding of device operating mechanisms^[9−13]. Compared to OLEDs, QLEDs offer superior performance, including lower costs, easier solution processing, and the ability to be manufactured at lower temperatures, making them highly competitive among light-emitting devices^[14−16]. Theoretically, QLEDs can achieve exceptionally high color purity up to 140% of the National Television Systems Committee's standard^[17].

Red and green QLEDs represent a new generation of light-emitting technology, demonstrating significantly enhanced performance. Although blue QLEDs have attracted considerable attention for their potential in display technology, they exhibit substantial drawbacks in efficiency and lifespan compared to their red and green counterparts. For example, red- and green-emitting indium phosphide (InP) QDs have achieved notable advancements, with external quantum efficiencies (EQEs) of 22.56%^[18] and 16.3%^[19], respectively. In contrast, blue-emitting InP QLEDs have EQE_max of only 2.8%^[20], limiting their broader application in full-color display technologies. The reason for this discrepancy can be analyzed in terms of both the blue QDs themselves and the device structure. Blue QDs have a larger bandgap, which increases the energy loss pathway for exciton quenching. In particular, the blue-shifted emission due to quantum confinement effect (QCE) makes it difficult to produce deep blue emission for certain kinds of QDs. In terms of synthesis, complex and precise tuning of the growth kinetics of nucleation, reaction time, use of multiple ligands and additional injection of precursors is required^{[21, 22]}. In terms of device structure design, blue QDs have a large bandgap and a deep VBM, leading to a large hole injection barrier at the interface between commonly used hole transport layers (HTLs) and blue QDs. Efficient hole injection cannot be achieved using conventional organic HTLs. However, electrons can be injected into the emission layer (EML) rapidly due to the use of inorganic electron transport layers (ETLs) (e.g., ZnO, ZnMgO)^[8]. To overcome these challenges, researchers are exploring novel QDs materials and device architectures, including core/shell-structured QDs to improve stability and optical efficiency and innovative charge transport layers (CTLs) to enhance charge injection and transport.

Over the past three decades, cadmium selenide (CdSe) QDs have been extensively studied due to their excellent PLQY and narrow full width at half maximum (FWHM). These properties enable Cd-based QLEDs to achieve high efficiencies, with EQEs exceeding 20%^[23]. However, the toxicity of cadmium (Cd) has impeded its commercial use and raised environmental concerns. In response, researchers are developing heavy-metal-free QDs, such as group Ⅱ−Ⅵ QDs (e.g., zinc selenide (ZnSe-based) QDs), group Ⅲ−Ⅴ QDs (e.g., InP, Gallium nitride (GaN) QDs), and carbon dots (CDs). While InP and ZnSe-based QDs show promise for blue light emission, achieving deep blue emission below 470 nm remains challenging. CDs are attractive for bioimaging due to their low cost, minimal toxicity, good biocompatibility, and tunable fluorescence properties, although their aggregation-induced quenching (ACQ) phenomenon in the solid state limits their use in electroluminescent (EL) devices^[24]. GaN has been extensively studied for its unique properties, yet research on GaN QLEDs remains limited, with most previous work concentrating on applications such as photodetectors and photocatalysts^{[25, 26]}.

This review summarizes the progress in blue QLEDs, covering Ⅱ−Ⅵ QDs, Ⅲ−Ⅴ QDs, and CDs. Section 2 introduces the basic properties of QDs to aid in understanding QDs synthesis and performance optimization and the Section 3 introduces the development history of QDs and QLEDs. Section 4 discusses CQDs synthesis methods and the nucleation and growth mechanisms. Section 5 details research advancements in enhancing blue QLEDs performance, including QDs materials, synthesis optimization, and device improvements. Finally, we will offer insights into the prospects and challenges of blue QLEDs and analyze future development trends. We aim to help researchers better understand and advance blue QLEDs performance, accelerating their application in solid-state lighting and flat-panel displays.

Since the size of CQDs typically ranges from 1 to 10 nm, approaching twice the exciton Bohr radius of their bulk counterparts, their electronic and optical properties change significantly with size. Notable changes include the discretization of energy levels and an increase in the band gap energy, both direct manifestations of the quantum size effect (QSE). As the size of the CQDs decreases further, the behavior of electrons and holes becomes more constrained, leading to energy level splitting and an increased energy gap between the valence and conduction bands (Fig. 1(a)). This is a concrete manifestation of the effect of QCE on the photoelectric performance of QDs which was first proposed by Kubo et al. in 1961, significantly impacting the synthesis and application of QDs^{[27, 28]}. Additionally, varying the size of CQDs can adjust their emission peaks (Fig. 1(b)). The low-dimensional structure of CQDs facilitates efficient electron−hole recombination in a confined space, enhancing luminescence across a broad spectrum from visible to near-infrared wavelengths via QCE.

The optical properties of QDs are influenced by surface effects and quantum tunneling. Due to their small size, a larger proportion of surface atoms is exposed, which can lead to surface reconfiguration and the introduction of forbidden in-band energy levels^{[29, 30]}. These surface states, often due to unsaturated chemical bonds, can be mitigated through chemical passivation, thereby enhancing the electronic and optical properties of QDs. Quantum tunneling effects, where electrons and holes tunnel through the QDs' finite potential well, impact energy level distribution and hopping behavior^[31]. As QDs size decreases, the tunneling effect becomes more pronounced due to the increased depth of the potential well, raising the tunneling probability. Currently, semiconductor NCs from Ⅱ−Ⅵ (e.g., CdSe, ZnSe, ZnS, CdS, ZnSe(Te)) and Ⅲ−Ⅴ (e.g., InP, InAs) groups have been extensively studied (Fig. 1(d)).

QDs are usually NCs made from semiconductors, similar to isolated atoms and molecules. QDs were first discovered by Alexei Ekimov in the Soviet Union in the 1970s^{[34, 35]}. Their research involved CuCl and CdSe NCs embedded in a glass matrix^[36]. Subsequently, Alexander Efros elucidated the mechanism of action of the new properties of these NCs, mainly the restricted range of electronic activity in the nanoscale range^[37]. As a result of the American−Soviet Cold War at the time, which in turn affected scientific communication, the American scientist Louis Brus^{[29, 38]} succeeded on the other side of the world in synthesizing NCs, later known as CQDs, in a liquid medium. He observed that the spectrum of QDs changed with increasing reaction time, and therefore suspected that the color change of QDs was due to size change. Then he proposed a theory involving mass and dielectric polarization to explain the observed phenomena^[39]. This explanation mentions that the spatial confinement of charge carriers can change the energy spectra of three-dimensional confined electrons, holes, and excitons, leading to specific size dependence in the absorption and luminescence spectra of QDs^[40].

With further research, more QDs synthesis methods such as physical vapor deposition^[41], molecular beam epitaxy^[42], chemical vapor deposition^[43], and plasma injection^[44] have been developed to achieve high quality QDs. Existing QDs synthesis methods are mainly categorized into two main groups: physical vacuum methods and wet chemical methods. The synthesis of monodisperse QDs by chemical liquid phase method has attracted wide attention due to the need for large-scale industrial production. In 1993, Moungi Bawendi^[45] obtained QDs with less than 5% size change by hot-injection method, which greatly improved the synthesis quality of QDs. This landmark work laid the foundation for contemporary CQDs synthesis. In recognition of their pioneering work on QDs, Moungi Bawendi, Louis Brus, and Alexei Ekimov were awarded the 2023 Nobel Prize.

With increasing public concern for environmental protection, the use of heavy metals has been severely restricted due to their potential toxicity. In order to minimize the release of hazardous chemicals, some measures have been taken, such as reducing the content of Cd through compositional control or reducing its release through encapsulation technology. Encapsulation techniques such as solvent evaporation, ionic gelation, and supercritical fluid encapsulation provide physical barriers that isolate hazardous chemicals from the environment, thereby reducing direct exposure and uncontrolled release^[46]. For example, alginate beads encapsulated with magnetic nanoparticles can be utilized to remove Pb²⁺ from water^[47]. In addition, QLEDs are encapsulated using UV-curable resin glue, which improves their device performance and stability on the one hand, and helps to prevent leakage of hazardous chemicals on the other^[48]. However, the development of heavy-metal-free QDs for zero emission of toxic chemicals remains an urgent issue. Particularly in the field of display and lighting technologies, the development of Cd-based QLEDs has become an important research focus due to the potential environmental and health risks associated with them. As a result, researchers are actively developing heavy-metal-free QDs, and research areas include ZnSe-based, InP, CDs, and GaN QDs^{[20, 49−51]}. Although these new QDs face challenges in achieving color purity and efficiency comparable to Cd-based QDs, significant progress has been made in developing high-performance, heavy-metal-free QDs and their corresponding QLEDs devices. We will elaborate on this part in Section 5.

At the end of the 20th century, people were not optimistic about the prospect of QLEDs in the EL display field because its EQE at that time was <1% and its brightness was only in the order of 10². Until after 2000, QLEDs has fully absorbed the mature OLEDs structure optimization and working mechanism experience. QLEDs has been rapidly developed, the performance has been steadily improved, and has reached the civil requirements. It is worth noting that the innovation of QDs and CTL plays the most important role in the development of QLEDs technology^[52]. The device structure of QLEDs has gone through roughly four stages of development. At first, Alivisatos employs a polymer/QDs bilayer structure, with CdSe QDs acting as both the EML and the ETL. Due to the poor conductivity and low PLQY of the CdSe QDs solid-state films without shell layers, these devices can only achieve weak brightness and very low EQE (<0.01%)^{[53, 54]}. Subsequent coating of CdS shells on the surface of CdSe QDs improved the peak EQE of QLEDs to 0.22%. However, the polymer luminescence observed in EL spectra indicated poor exciton confinement in the QDs layer^[55]. In 2002, Coe et al.^[56] prepared type-Ⅱ structured QLEDs, which were similar to OLEDs, by utilizing organic materials as ETLs and HTLs, and the application of organic CTLs enhanced the peak EQE of the devices to 0.52%. In this type of QLEDs, the exciton formation is mainly dominated by Förster resonance energy transfer (FRET), which is quite different from direct charge injection^[57]. However, the relatively low conductivity of organic materials limits the injection of charge carriers, and the carrier leakage due to pinholes in monolayer QDs limits the efficiency of these devices. Replacing organic CTLs with inorganic CTLs to form a type Ⅲ structure improves the stability of the device in air and enables higher current density. The Bawendi group^[58] reported fully inorganic QLEDs with current densities up to 4 A∙cm⁻² in 2008, but the device EQE was less than 1%. The main reasons for the low device efficiency are the damage of QDs during sputtering, carrier injection imbalance, and PL burst of QDs by the surrounding conducting metal oxides^{[59, 60]}.

Since 2010, Type Ⅳ hybrid structures with inorganic ETLs and organic HTLs have been employed^[61−63]. Compared to Type Ⅲ structures, Type Ⅳ structures enable a more balanced and efficient carrier transport, which enhances the overall device efficiency. However, the injection rate of electrons in the Type Ⅳ structure is typically larger than that of holes, leading to an imbalance in carrier injection. To mitigate this, strategies such as limiting electron injection or boosting hole injection are commonly applied. In 2014, Peng Xiaogang's team^[64] inserted a polymethylmethacrylate (PMMA) insulating layer between QDs and ZnO NCs, which suppressed electron injection and prevented the bursting of excitons by the ETLs of the ZnO NCs, and realized for the first time a high-efficiency organic−inorganic hybrid-structured red QLEDs with an EQE of more than 20%. However, such high-efficiency devices require precise control of the insulating layer thickness of less than 10 nm, which poses a great difficulty for industrial production. Thus, in 2018, they prepared insulating-layer-free red QLEDs with Mg-doped ZnO NCs as the ETL^[65]. Compare to the insulating-layer-free device with ZnO NCs as the ETL, the Mg doping significantly suppressed the exciton bursting at the QDs/ETL interface. They then balanced the carrier injection by further adjusting the electrode thickness, and finally achieved the highest EQE of 18.1% for red insulating layer-free devices at that time. On the other hand, Shen et al.^[66] took the approach of lowering the hole injection barriers to enhance the device efficiency. They successfully adjusted the energy level of the EML to match the energy level of the HIL by cladding the QDs with a ZnSe shell layer. They realized red, green, and blue QLEDs devices with maximum EQEs of 21.6%, 22.9%, and 8.05% and peak brightnesses of 356 000, 614 000, and 62 600 cd∙m⁻², respectively.

In the 1950s, the LaMer model was introduced to explain the formation of colloidal particles through a monodisperse nucleation and growth process^{[45, 67]}. This model remains widely used in nanomaterial synthesis, particularly in the preparation of nanoparticles and QDs. The model divides the formation process into three distinct stages: monomer formation (Stage Ⅰ), nucleation (Stage Ⅱ), and growth (Stage Ⅲ). In Stage Ⅰ, precursors decompose to generate monomers, leading to a supersaturated solution. During Stage Ⅱ, nucleation occurs rapidly once the monomer concentration surpasses a critical threshold, initiating the formation of stable core. Finally, in Stage Ⅲ, particle growth occurs as monomers deposit onto existing core, while the nucleation process is suppressed due to the reduced monomer concentration. The LaMer model underscores the importance of controlling monomer concentration and reaction conditions to achieve a desired particle size and uniformity. Although the LaMer model does not fully account for complexities such as size distribution broadening or the effect of surface chemistry on particle formation and growth kinetics^[68], it provides key insights into the nucleation and growth mechanisms, which can be very helpful in understanding the nucleation and growth of CQDs.

High-quality QDs necessitate precise structural design and efficient synthesis to ensure commercial viability. Blue QDs, widely used in lighting and display technologies, are typically synthesized in organic phase systems. This approach helps circumvent common issues associated with aqueous phase systems, such as low crystallinity, uneven size distribution, low luminescence efficiency, and poor stability. The main methods for synthesizing QDs include the hot-injection method (Fig. 2(a)) and the heat-up method.

The hot-injection method is a widely used technique for synthesizing QDs. In this method, precursors are rapidly injected into a hot coordinating solvent (Fig. 2(b)), triggering a sharp supersaturation that results in a burst of nucleation. This rapid nucleation corresponds to the second phase of the LaMer model, where the monomer concentration exceeds the critical threshold, leading to uniform seed formation^[69]. Subsequent growth of QDs occurs as the monomer concentration decreases, preventing further nucleation and promoting controlled particle growth. It allows for precise control over the nucleation process and fine-tuning of both QDs size and optical properties. In the preparation of high-performance ZnSe-based QDs, injecting Zn and Se precursors-either sequentially or concurrently-into a high-temperature reaction system induces monomer supersaturation. This leads to rapid core formation due to the free energy in the solution. Subsequently, the decrease in monomer concentration inhibits new nucleation and promotes the growth of NCs. Hines et al. were the first to synthesize high-quality monodisperse colloidal ZnSe NCs using Et₂Zn and HDA/TOP as the Zn source and solvent via hot injection, setting a precedent for ZnSe(Te) QDs^[70]. However, Et₂Zn, due to its toxicity and flammability, has been replaced by more environmentally friendly alternatives such as zinc stearate (Zn(St)₂)^{[71, 72]}, zinc acetate (Zn(OAc)₂)^[73], ZnO^[74], and zinc carboxylate salts^[75]. For instance, Ippen et al. used diphenylphosphine-selenium (DPP-Se) injected into excess Zn precursor to form ZnSe cores, with transmission electron microscopy (TEM) images demonstrating the resulting particle size variations (Fig. 2(d))^[75]. Organic phosphines like trioctylphosphine (TOP) and DPP serve as both ligands and solvents in these processes, coordinating with non-metallic elements (Se and Te) and influencing reaction kinetics crucial for ZnSe(Te) QDs^[68].

The synthesis of InP QDs faces significant challenges due to the high reactivity of atomic species and substantial reaction barriers. To overcome these difficulties, researchers have utilized tris(trimethylsilyl)phosphate ((SiMe₃)₃P) as a highly reactive phosphorus precursor. In 1994, Nozik et al. pioneered the synthesis of InP QDs by combining InCl₃ with (SiMe₃)₃P^[76]. Subsequently, Peng et al. accelerated the nucleation process from days to hours by employing octadecene (ODE) as a non-coordinating solvent and introducing fatty acids as ligands^[21]. Moreover, Choi et al. advanced the field by employing asymmetric In-carboxylate and Ga-carboxylate complexes to synthesize InGaP alloyed QDs, while symmetric complexes were utilized to produce core/shell (C/S) InP/GaP QDs (Fig. 2(e))^[77]. These studies provide a critical theoretical and experimental foundation for the efficient synthesis of InP QDs.

Another commonly used synthesis strategy is the heat-up method, which involves either mixing pre-prepared cationic and anionic solutions or dissolving and adjusting the temperature after combining all raw materials. This approach results in a prolonged nucleation stage by gradually heating the precursors to induce monomer formation. Although this method is slower, it allows for continuous nucleation and growth under milder conditions^[78]. For ZnSe-based QDs, the heat-up method simplifies the synthesis by avoiding high-temperature injections and mixing precursors and ligands at lower temperatures^[68]. With the coating of ZnSe and ZnS shell layers, the size of ZnSe(Te) QDs gradually increases and the PLQY is enhanced from 23% to 63%, and the shape and size of the QDs also become more homogeneous as can be seen from the TEM images (Fig. 2(c))^[79]. Additionally, hydrothermal^[80] and microwave-assisted^[81] methods have been employed for synthesizing ZnSe-based QDs. These methods are environmentally friendly and suitable for large-scale production, with potential for creating water-soluble QDs. However, they often suffer from issues such as poor stability, low PL intensity, and broad emission line widths compared to organic-phase methods, which can limit their applications in lighting and display technologies.

Notably, blue light-emitting QDs are characterized by their wide bandgap, which is the primary factor responsible for their blue emission. Generally, because of relatively small core volumes, blue-emitting QDs tend to have large specific surface areas and a high number of exposed surface atoms, which can result in numerous unsaturated dangling bonds and surface states with varying energy levels. These surface defects can significantly influence the optoelectronic performance of the QDs. Epitaxial growth of other semiconductor materials (e.g., CdS, ZnS, ZnSe QDs) on the QDs core to form C/S structures can effectively passivate both anionic and cationic surface sites, thereby enhancing the QDs' optoelectronic performance. Furthermore, surface ligands serve a dual role in QDs. In the case of long-chain organic ligands like oleic acid (OA) and oleylamine (OAm) in C/S QDs they serve as protective agents during QDs synthesis. However, these lengthy ligands can impede the transport of charges from the transport layer into the QDs, leading to a decline in device performance. Therefore, reconstructing ligands on the QDs surface is also essential.

ZnSe, a semiconductor QD with a direct bandgap of 2.7 eV, is gaining recognition as a promising alternative to Cd-based materials for blue-light applications, and is regarded as one of the leading heavy-metal-free blue-light emitters^{[84, 85]}. Although research on ZnSe-based QDs began in 1998^[70], their development has been slower compared to Cd and Cd-based QDs, mainly due to inherent defects. The surface of ZnSe QDs is susceptible to oxidation or hydroxylation, leading to localized trap states in the bandgap. This issue is more severe in ZnSe(Te) alloy QDs because Te is more reactive to oxygen than Se, making ZnSe(Te) cores highly unstable and requiring an oxygen-free environment during synthesis^[68]. These challenges can be mitigated through functional modifications, mainly focusing on two approaches: surface passivation and bandgap engineering^[68].

Surface passivation encompasses three key approaches: shell passivation, chemical etching, and ligand modification. Shell passivation entails the epitaxial growth of additional semiconductor layers on the ZnSe(Te) core, creating a C/S structure. This approach passivates cationic and anionic surface sites, shields the material from degradation caused by moisture and oxygen exposure, and minimizes electron and hole trapping at defect sites, leading to improved PLQY and enhanced stability. ZnS is a commonly used shell material that improves PLQY when used to form ZnSe/ZnS C/S QDs. Gao et al. demonstrated this by synthesizing ZnSe cores smaller than 10 nm, adding ZnSe shell layers, and coating them with approximately four monolayers of ZnS, resulting in ZnSe/ZnS C/S QDs with a PLQY of up to 95% and narrow emission widths (<9.6 nm) (Fig. 3(a))^[84]. Chemical etching is another prevalent surface treatment, particularly for ZnSe(Te) cores prone to oxidation. Treatment with chemicals like hydrofluoric acid (HF) or zinc chloride (ZnCl₂) removes oxides and other defects from the surface, enhancing QDs stability and luminescence efficiency (Fig. 3(b)). For instance, Kim et al. utilized HF and ZnCl₂ to nearly eliminate stacking errors in the ZnSe(Te) crystal structure, achieving a PLQY of 93%^[49]. They also provided a detailed synthetic scheme for ZnSe(Te) cores, ZnSe(Te)/ZnSe C/S, and ZnSe(Te)/ZnSe/ZnS C/S/S QDs, including TEM images (Fig. 3(c)). Ligands are crucial for the optical properties and stability of QDs. Ligand engineering, including the passivation of inorganic halide ions and modification of short-chain ligands, improves QDs efficiency and stability. For example, Park et al. significantly enhanced the PLQY and stability of ZnSe(Te)/ZnSe/ZnS QDs under UV irradiation through halide passivation^[86]. Additionally, replacing OA with 1-dodecanethiol (DDT) in ligand exchange strategies further improves PLQY stability under ambient conditions.

Blue-emitting ZnSe(Te) QDs have been widely studied, demonstrating remarkable properties such as PLQY reaching 100% and FWHM as narrow as <10 nm. These characteristics make them strong contenders to replace Cd and Cd-based QDs. However, obtaining pure blue emission from ZnSe-based QDs remains a challenge. While Te doping can tune the bandgap, it also introduces defects, leading to lower efficiency and spectral broadening^[87−90]. Additionally, ZnSe(Te) QDs encounter stability issues. Bandgap engineering, through size control and alloying, offers an effective method for adjusting QDs emission wavelengths. The bandgap of ZnSe(Te) NCs inversely relates to their radius, causing red-shifts in absorption and PL peaks as size increases. Controlling parameters such as reaction temperature, precursor concentration, Zn/Se ratio, and heating rate allows manipulation of ZnSe(Te) QDs size, shape, and crystal phase, thereby fine-tuning their optical properties. For example, Yang’s group successfully expanded ZnSe(Te) QDs size from 5.3 to 12.2 nm by sequentially introducing Zn and Se stock solutions^[91]. They added a ZnSe inner shell to the ZnSe(Te) core, with the size and shell thickness of three ZnSe(Te) QDs detailed in Fig. 3(d). Doping with Te forms a ZnSe(Te) ternary alloy, which enables broad emission wavelength tuning from violet to green by adjusting the Te/Se ratio. Differences in ionic radii between Te and Se can cause lattice mismatches and interfacial defects, affecting PLQY and spectral linewidth. Additionally, bandgap variations in ZnSe(Te) alloyed QDs, known as band bending, result from structural relaxation effects due to differences in lattice constants and bond lengths. As Te doping increases, PLQY tends to decrease, showing PL tailing and spectral broadening in the low-energy region. Redshifting of the PL spectrum, broadening of the half-peak width, enhancement of the tail emission and asymmetric spectra seem to be universal as the Te/Se ratio increases^{[49, 88, 92]}. Lee et al. suggested that during ZnSe(Te) alloy core synthesis, interdiffusion between the pre-grown ZnSe(Te) core and ZnSe shells likely occurs as shell thickness increases^[92]. This interdiffusion expands the effective domains of the ZnSe(Te) core, reducing the relative Te content, which mitigates tail emission and improves spectral symmetry towards narrower emission or higher color purity (Fig. 3(e)).

Low-toxicity, high PLQY, and color-tunable ZnSe(Te) QDs were successfully synthesized, establishing a solid foundation for advancing high-performance QLEDs technology. ZnSe EL devices were scarcely studied until 2020, with few reports and EQEs below 10% in almost all binary ZnSe systems. Additionally, the wide bandgap of ZnSe causes the EL peaks of these devices to predominantly fall within the UV region, specifically 420 to 440 nm^[68]. Over the past three years, various strategies have yielded QLEDs with emission peaks in the blue region, significantly enhancing device efficiency. In the last decade, the EQE of blue QLEDs has risen from 0.65% to 20.2%.

As early as 2012, Xiang et al. developed the first QLEDs using ZnSe/ZnS C/S QDs as emitters, achieving an EQE of 0.65% at a wavelength of 420 nm^[93]. This result was obtained by carefully selecting the HTL material, optimizing its thickness, adjusting the EML thickness, and improving charge balance. Ji et al. achieved a highly narrow FWHM of approximately 17.2 nm, with exceptional color purity in blue EL, by utilizing Poly(N,N-bis(4-butylphenyl)-N,N-bis(phenyl)benzidine (Poly-TPD) and 2,2',2''-(1,3,5-benzenetriyl)tris(N,N-diphenylamine) (2-TPD) as HTLs through an energy transfer mechanism^[94]. To improve the ETL, Han et al. introduced a Mg(OH)₂ layer on surface-modified ZnMgO (m-ZnMgO) nanoparticles, likely formed through a secondary magnesium reaction on the ZnMgO surface^[95]. The Mg(OH)₂ layer reduced electron mobility, helping to mitigate the luminescence burst phenomenon in the EML and enhancing both charge balance and the energy band alignment of blue QLEDs, as depicted in Fig. 3(f). These advancements resulted in the creation of efficient blue QLEDs with a maximum luminance (L_max) of 2904 cd∙m⁻² and an EQE_max of 9.5%. More recently, Kim et al. fabricated a ZnSe(Te) QLEDs with an EQE_max of 20.2%, marking it as the most efficient blue ZnSe(Te) QLEDs to date, with a L_max of 88 900 cd∙m⁻² and a T₅₀ (the time required for brightness to drop to 50% of its initial value) of 15 850 h^[49]. Therefore, ZnSe(Te) QDs show great potential as promising candidates for next-generation, eco-friendly blue light sources and display technologies.

InP QDs are regarded as an environmentally friendly nanomaterial and have attracted significant interest due to their low toxicity, high extinction coefficient, large Bohr exciton radius, and wide emission wavelength range. With a bulk band gap of 1.35 eV and an exciton Bohr radius of 10 nm, InP QDs exhibit a significantly larger exciton radius compared to CdSe QDs, which have a Bohr radius of 4.6 nm. This increased exciton radius presents a distinct intrinsic advantage^{[21, 96]}. The larger exciton Bohr radius of InP QDs offers several intrinsic advantages. First, it enables stronger QCE, allowing for precise modulation of emission wavelengths through size control, which broadens their applicability in displays and lighting technologies^[21]. Second, a larger exciton radius facilitates a broader absorption spectrum, leading to better light-harvesting efficiency in optoelectronic devices^[97]. Third, it enhances exciton mobility and reduces the recombination rate of charge carriers, which are beneficial for improving device efficiency and stability^{[98, 99]}. Lastly, the increased exciton radius contributes to higher PLQY due to reduced Auger recombination (AR) and improved energy transfer processes, making InP QDs an excellent alternative to Cd-based QDs for environmentally friendly applications. However, the performance of blue InP QLEDs is significantly inferior to that of red and green QDs. This discrepancy arises because their smaller core size is more prone to defects and issues related to valence and lattice mismatch between the core and shell^{[100, 101]}. Among these issues, valence mismatch strongly affects the EL performance of the device. It results in non-radiative recombination during electron and hole injection, as well as carrier capture by defects, thereby limiting device performance. Specifically, the small potential barrier between the conduction band minimum and the lowest unoccupied molecular orbital (LUMO) energy level of the HTL in blue InP QDs facilitates electron transfer across the barrier into the HTL, where electrons do not effectively recombine with holes in the QDs layer. This results in a device with high current density but low brightness and efficiency^{[18, 102]}.

Since the first synthesis of InP QDs in 1989, extensive research has focused on their emission spectra, ranging from blue to near-infrared^[103−105]. Early QDs lacked protective shell layers due to the direct bonding of surface ligands, which resulted in numerous surface defects, broad and asymmetric emission spectra, and low luminescence efficiency^{[21, 106, 107]}. The broad-spectrum luminescence observed was not an intrinsic property of InP but rather a result of inhomogeneous size distribution. To address these issues, researchers have employed various strategies, including introducing cations during nucleation, exploring lattice-matched shell materials, and developing C/S structures with reduced defects^[108−111]. Additionally, the easily oxidizable nature of InP contributes to defect formation, making surface etching an effective method for defect reduction^{[112, 113]}. Energy level mismatches between core and shell materials also lead to severe electron delocalization, negatively impacting PLQY and leading to nonradiative losses such as FRET and AR^[114]. To mitigate these issues, researchers have proposed various design strategies for C/S structures, including the addition of intermediate shell layers and optimizing lattice compatibility at the C/S interface^[115−117].

Elemental doping at the C/S interface can have dual effects depending on the doping strategy and conditions. On one hand, improper doping may exacerbate surface defects by introducing nonradiative recombination centers. On the other hand, carefully optimized doping can mitigate these interface defects by enhancing lattice compatibility, passivating surface states, and improving the chemical composition at the interface. For example, metal ions such as Ga³⁺, Eu³⁺, Cd²⁺, Zn²⁺, Mn²⁺, Cu⁺, and Ag⁺ have been employed during the synthesis of InP QDs to improve performance^[118−125]. Among these, Zn²⁺ and Ga³⁺ doping have shown particular effectiveness in passivating defects and enhancing luminescence by reducing nonradiative recombination. The large lattice constants of InP QDs also facilitate alloying or surface doping to support the growth of subsequent shell layers. Despite these advancements, the EQE of blue InP QLEDs remains below 3%^[20], reflecting the increasing difficulty of synthesis from red to green to blue. This analysis reviews cases of achieving high-quality EL to offer valuable insights for the further development of EL devices.

For instance, Shen et al.^[126] synthesized InP cores with uniform size by controlling the reaction rate, using the cost-effective tris(dimethylamino)phosphine ((DMA)₃P) as the phosphorus precursor. The fluorescence properties and stability of the QDs were enhanced by applying a multilayer ZnS shell using a zinc-halogen-mediated colloidal technique. Additionally, the InP QDs size was adjusted by modifying the ratio of halogen ligands (X = Cl, Br, I) to indium ions (In³⁺) to promote blue light emission. By optimizing the P/In and I/In ratios, they significantly boosted the PLQY of the InP QDs, reaching up to 76%. These QDs demonstrated excellent stability, maintaining performance for over 1000 h under ambient conditions (Fig. 4(b)). For device application, ZnMgO was used as the ETL, resulting in QLEDs that emitted blue light with a L_max of 90 cd∙m⁻² at 10 V. In 2020, Zhang et al. introduced a thin GaP bridging layer between the core and shell via a shell engineering approach, effectively reducing the lattice mismatch between the InP core and ZnS shell (Fig. 4(c))^[127]. They fabricated blue QLEDs using QDs with both thick and thin shells in the EML, finding that the thick-shell device exhibited superior performance, achieving a high luminance of L_max = 3120 cd∙m⁻² and a EQE_max = 1.01%.

The Sun group similarly utilized (DMA)₃P as a phosphorus source to synthesize InP cores by injecting precursors of ZnI₂ and InI₃ at 200 °C^[128]. ZnS was chosen as the shell material, resulting in the successful development of blue InP/ZnS/ZnS QDs with an emission wavelength of 468 nm and a PLQY of up to 45%. This was accomplished by using an excess of zinc stearate and introducing TOP-S during ZnS shell formation, which effectively eliminated residual zinc stearate, increased the shell thickness, and minimized energy transfer between QDs, thereby enhancing both stability and PLQY. Furthermore, they explored the use of these novel QDs in QLEDs, significantly improving the EQE from 0.6% to 1.7% by optimizing carrier injection and increasing the shell layer thickness.

In a subsequent study, the Sun group^[129] employed (DMA)₃P as a phosphorus precursor and replaced ZnI₂ with ZnBr₂ for the ZnS shell layer. This modification enhanced the surface stability of the QDs by passivating the surface with bromide ions (Br⁻), which reduced core electron diffusion into the shell layer and consequently minimized nonradiative recombination. The improved passivation also more effectively addressed surface defects (Fig. 4(a)). Additionally, increasing the ZnS shell layer thickness limited core electron diffusion, further enhancing the PLQY. Degassing treatment after encapsulation increased the quantum yield of the InP/ZnS QDs by approximately 10%, achieving a PLQY of 93%. The ligand exchange was performed by replacing the long-chain DDT with the short-chain 1-octanethiol (OT), which proved to be an effective method for improving carrier injection efficiency. By utilizing this method in conjunction with 2,2',7,7'-tetrakis(N,N-diphenylaminephenyl)9,9'-spirobifluorene (TFB) as the HTL and Zn_xMg_1−xO as the ETL, the EQE of InP QLEDs improved from 1.8% to 2.6% (Fig. 4(d)). Additionally, the L_max of the QLEDs increased from 174 to 422 cd∙m⁻², while the current density at 6 V rose from 46 to 101 mA∙cm⁻².

Kim et al. introduced a Ga-doping method to overcome the challenges associated with the EL of pure blue QDs^[130]. Pre-synthesized InP QDs can be successfully alloyed with Ga at 280 °C using GaI₃, with the degree of alloying controlled by adjusting the amount of GaI₃ (Fig. 4(e)). By applying sequential surface passivation with ZnSeS inner shells and ZnS outer shells, InGaP/ZnSeS/ZnS QDs are produced, offering tunable blue emission between 465−475 nm, based on the Ga content. QDs with a 465 nm emission wavelength achieve a PLQY of up to 80%. The resulting QLEDs show outstanding performance, with a L_max of 1038 cd∙m⁻² and an EQE of 2.5%. These advancements in ternary InGaP QDs present a promising path toward high-quality blue QLEDs.

CDs are a type of carbon-based nanomaterial characterized by at least one dimension measuring less than 10 nm. They were first identified by Xu et al. in 2004 while isolating and purifying single-walled carbon nanotubes (SWCNTs)^[131], CDs are notable for their low toxicity, biocompatibility, environmental friendliness, low cost, chemical stability, and ease of synthesis. These attributes position CDs as a promising alternative to conventional semiconductor QDs. CDs offer tunable fluorescence properties and photostability, and their surfaces can be passivated and functionalized to further tailor their physicochemical properties. These advantages make CDs particularly attractive for optical applications, environmentally friendly preparation, cost-effectiveness, and biological uses, garnering considerable attention from the research community.

Due to the diverse raw materials used in the preparation of CDs, developing a unified theory from the literature results is challenging. The synthesis of CDs primarily falls into two categories: top−down and bottom−up approaches. Top−down methods involve deconstructing larger carbon structures and have been instrumental in the discovery and characterization of CDs. In contrast, bottom−up methods involve synthesizing CDs from molecular precursors through processes such as combustion or heat treatment. For instance, Yuan et al. synthesized multicolor bandgap fluorescent carbon dots (MCBF-CDs) with emission ranging from blue to red by controlling the fusion and carbonation reactions of citric acid (CA) and diaminonaphthalene (DAN), and by doping with nitrogen^[132]. These CDs exhibit homogeneous atomic arrangements and a high degree of crystallinity, demonstrating strong exciton absorption bands and minimal overlap between exciton absorption and emission spectra, which enhances fluorescence efficiency, achieving a PLQY of 75% for the blue CDs. To investigate the EL properties of MCBF-CDs, LEDs with the structure ITO/PEDOT/MCBF-CDs/TPBi/Ca/Al were fabricated without conventional HTLs and the blue devices exhibited a L_max of 136 cd∙m⁻².

CDs synthesized from CA as a precursor face the challenge of spectral broadening. Sargent’s team used density functional theory (DFT) to investigate the influence of oxygen-containing functional groups on this phenomenon. Their findings revealed that the carboxyl (COOH) group induces a strong polarization effect, which alters the degree of wave function localization through relative plane rotation, ultimately resulting in spectral broadening^[50]. To address this problem, they employed an effective edge amination strategy to eliminate oxygen-containing functional groups, reducing surface defects and trap states. The researchers developed a two-step method for the synthesis of high-colour-purity deep-blue CDs (HCP-DB-CDs) (Fig. 5(a)). These CDs exhibit tunable fluorescence emission across a range from deep blue to near-infrared regions. Additionally, they emit light with a narrow FWHM of less than 35 nm in the deep blue region and achieve a PLQY of up to 70% ± 10% (Fig. 5(b)). LEDs were fabricated by optimizing the concentration of CDs in EML using TFB as the HTL and poly(N-vinylcarbazole) (PVK) as the host material for CDs, demonstrating high brightness L_max over 5240 cd∙m⁻² and high EQE of 4% (Fig. 5(c)).

Zhang et al. synthesized CDs using a one-step hydrothermal method with maleic urea and CA as precursors. The CDs were then treated with 30% hydrogen peroxide (H₂O₂) to remove amino functional groups, resulting in a chemically inert surface. This treatment preserved the fluorescence properties of the CDs, demonstrating excellent stability^[133]. The PL peak of the CDs in ultrapure water was observed at 446 nm with a PLQY of 26.4%. For device applications, the CDs were doped into PVK as the active EML, and lithium 8-hydroxyquinolines (Liq) were used as the EIL. By adjusting the mass ratio of CDs to PVK, the 40 wt% CDs-PVK mixture exhibited optimal luminescence efficiency and low driving voltage in the LEDs. TRPL analysis confirmed energy transfer between PVK and CDs via the FRET mechanism. The CDs-based LEDs achieved a L_max of 223 cd∙m⁻² with an EQE of 0.856%. The unencapsulated CDs-LEDs maintained high stability, with lifetimes exceeding 217 h when brightness dropped to 65%.

To address the issue of ACQ in CDs, An et al. synthesized CDs with multicolored solid-state fluorescence properties using CA, urea, and phenylethylamine as precursors^[134]. Phenylethylamine acted as a co-carbonization agent, introducing a phenylethyl structure that formed a protective shell, thereby effectively preventing fluorescence quenching in the solid state. Similarly, Zhang et al. enhanced the solid-state fluorescence properties of CDs by introducing hydrogen bonds and polymer chains^[135]. They synthesized red, green, and blue CDs using a solvothermal method with o-phenylenediamine, urea, and phenylethylamine as precursors, varying reaction conditions and solvents. The blue CDs achieved a PLQY of up to 59.75%. Additionally, a co-host system containing 4-[1-[4-[4-[bis(4-methylphenyl)amino]phenyl]cyclohexyl]-N-(3-methylphenyl)-N-(4-methylphenyl)aniline (TAPC) was designed by optimizing the doping concentration of CDs in PVK. This optimization significantly reduced the hole injection barrier and limited electron injection, thus balancing carrier injection and addressing the fluorescence burst issue in solid-state CDs (Fig. 5(d)). The resulting CDs-based LEDs achieved L_max of 827.60 cd∙m⁻².

GaN QDs are CQDs with direct bandgaps and two crystal structures: hexagonal wurtzite (α-GaN) and cubic zinc-blende (β-GaN). The bandgap of α-GaN is 3.4 eV, while that of β-GaN is 3.2 eV. GaN bulk semiconductors are widely studied for their diverse optoelectronic applications, such as blue LEDs. In 1996, Xie et al. first reported a low-temperature (280 °C) solution synthesis method for α-GaN NCs (30 nm in diameter), which provided a foundation for subsequent low-temperature synthesis techniques^[137]. In 2011, Dimos et al. developed a low-temperature (650 °C) synthesis method for β-GaN QDs using an ammonia gas stream without an organic gallium precursor. They achieved uniform QDs sizes at lower temperatures by employing ordered mesoporous SiO₂ (MCM-41) as a host matrix, which restricted the growth dimensions and sizes of the GaN QDs. Additionally, they synthesized α-GaN QDs at 365 °C through ammonia decomposition using the organic precursor tris(dimethylamido)gallium(Ⅲ)^[138].

Choi's group successfully synthesized α-GaN QDs using a low-temperature hot-injection method with lithium hexamethylsilylideneammonium (LiHMDS) as the nitrogen precursor, combined with GaCl₃ and stearic acid. They achieved wavelength tuning from the deep ultraviolet to the visible region by doping with Zn^[136] (Fig. 5(e)). Building on this work, Li et al. optimized the synthesis further by incorporating ZnCl₂ and developing a multiple washing post-treatment technique^[51]. Utilizing mixed solvents of chloroform and ethanol for size screening and morphology optimization improved film continuity and reduced size distribution inhomogeneity, thus laying a strong foundation for device fabrication. For the device structure, they designed a novel QLEDs configuration, comprising an ITO anode, ZnO ETL, QDs EML, electron-blocking layer (CBP), hole-blocking layer (MoO₃), and Al cathode. The fabricated QLEDs demonstrated typical diode characteristics, as shown by the J−V curve (Fig. 5(f)). Ultimately, the GaN QDs-based QLEDs achieved an EQE of 0.004% (Fig. 5(g)). This accomplishment represents the first application of GaN QDs in QLEDs and offers valuable insights for the future development of high-performance QLEDs.

CQDs have attracted widespread interest because of their exceptional optoelectronic characteristics, such as high PLQY, narrow emission spectra, excellent color purity, stability, and ease of processing in solution. QLEDs technologies leveraging these CQDs exhibit considerable potential in light-emitting devices due to their low cost, solution processability, and compatibility with low-temperature fabrication. In terms of QLEDs device preparation, the article describes the evolution from primitive polymer/QDs bilayer structures to organic−inorganic hybrid structures, which significantly improve the device performance by precisely controlling the injection of electrons and holes. In particular, higher efficiency and longer lifetime QLEDs have been achieved through the introduction of C/S structured QDs and optimized CTLs. While significant progress has been made with red and green devices, developing efficient blue QLEDs continues to pose a considerable difficulty. This paper reviews the progress in research on group Ⅱ−Ⅵ (ZnSe(Te)), Ⅲ−Ⅴ (InP, GaN) QDs, and CDs for blue QLEDs applications.

ZnSe-based QDs are promising candidates for blue light emission due to their large bandgap. Through size tuning and Te doping, ZnSe QDs can achieve emission across a broad spectrum from violet to green. However, Te doping introduces challenges such as FWHM broadening and spectral asymmetry, which must be addressed to enhance the efficiency and color purity of blue ZnSe-based QLEDs. InP-based QDs struggle with core sizes less than 2 nm, which, combined with inter-core-shell valence and lattice mismatch issues, limits device performance. Currently, the maximum efficiency of blue QLEDs using InP-based QDs is only 2.8%. Improving luminescence performance by reducing C/S interface defects through elemental doping is a promising approach. For instance, InGaP QLEDs have achieved a luminescence wavelength of 469 nm and an EQE of 2.5%, indicating that InGaP QDs may be a key future development for InP-based blue QLEDs. The ACQ effect greatly restricts the luminescence efficiency of CDs in their solid state. This challenge can only be addressed by synthesizing anti-burst CDs and embedding them in host materials. Although LEDs based on CDs have not yet reached the performance levels of inorganic QLEDs, their non-toxicity and biocompatibility make them promising candidates for bioimaging applications. GaN QDs have been less extensively studied, primarily due to the challenges associated with suitable low-temperature synthesis methods. Preliminary reports on GaN QLEDs are emerging, and it is anticipated that further research in this area will expand.

Improving efficiency and color purity will remain central to advancing blue QLEDs in the future. Research will prioritize refining C/S structures and optimizing charge transport materials to achieve greater brightness and longer device lifetimes, especially in blue QLEDs. Innovations in structure and materials, such as InP and ZnSe, can enhance luminescence and stability without heavy metals, supporting environmentally friendly applications. Advanced engineering of QDs surfaces and the development of new shell materials are expected to further boost performance.

Flexible substrates will unlock new opportunities for QLEDs in wearable and foldable displays as materials improve. Integrating flexible substrates like polymers or oxide-based films enables QLEDs to withstand bending and folding without compromising performance. The focus is on not only the robustness of QDs but also on creating flexible packaging and device layers that can be processed at low temperatures for stability. Flexible QLEDs are especially promising for next-generation consumer electronics, including foldable smartphones, rollable displays, and wearable devices.

Large-area QLEDs manufacturing is also crucial for scalable display and lighting applications. Key challenges include achieving uniform QDs layers over large surfaces and ensuring color consistency across extensive areas. Emerging large-area fabrication techniques, such as advanced inkjet printing, spin coating, and spray coating, demonstrate significant potential to enhance the deposition uniformity and production scalability of QLEDs on wafer-scale or flexible substrates. Paired with flexible substrate technology, these advancements will enable expanded, lightweight QLEDs panels for immersive home theater systems, digital signage, and even architectural lighting.

In summary, improvements in QLEDs efficiency, flexible form factors, and large-area capabilities will transform the display and lighting industries, enabling high-performance, eco-friendly devices for today’s most demanding applications.