1 Introduction
Metal halide perovskites have emerged as a revolutionary class of semiconductor materials for optoelectronic applications, owing to their exceptional photophysical properties, including high photoluminescence quantum yields (PLQYs) [
1], tunable band gaps [
2], and process compatibility [
3,
4]. In recent years, perovskite light-emitting diodes (PeLEDs) have rapidly advanced as a promising alternative to conventional organic LEDs (OLEDs) and quantum-dot LEDs (QLEDs). They have demonstrated remarkable progress in external quantum efficiency (EQE), which has surged from below 1% to over 30% within a decade [
5,
6]. Notably, perovskites offer precise and convenient color tuning without requiring molecular structural modifications, unlike OLEDs. Their emission spectrum spans the entire visible range, with the narrow full-width at half maximum (FWHM) matching the REC.2020 color gamut requirements for ultra-high-definition displays, demonstrating superior color purity compared to organic light-emitting materials. In 2020, our team compiled a performance table for PeLEDs, providing researchers with a valuable resource [
7]. PeLEDs have also achieved significant efficiency gains in recent years. These milestone breakthrough positions PeLEDs to compete with commercial OLEDs [
8]. Beyond outstanding device performance, perovskite materials offer substantial advantages in low-cost manufacturing [
9]. Unlike organic materials requiring complex synthesis or inorganic quantum dots, perovskite reactions are remarkably simple, achievable within an hour or even minutes using straightforward methods like one-step spin coating or the antisolvent process. Their synthesis typically involves basic metal halide precursors and minimal additives, further reducing PeLEDs manufacturing costs. Encouraged by these advantages, an increasing number of scientific reports are now dedicated to enhancing the efficiency and stability of PeLEDs, as well as exploring integration strategies, laying the groundwork for the commercialization of PeLEDs display technology [
10–
12].
However, several critical and unavoidable challenges remain in the commercialization of PeLEDs. First, regarding performance, both red and green PeLEDs achieve an EQE exceeding 30%, meeting a key commercial requirement. The EQE of blue PeLEDs has yet to surpass 30%, posing significant obstacles to full-color perovskite displays [
13,
14]. Second, active-matrix displays require patterning technologies [
15–
17]. Currently, most reported high-efficiency, stable perovskite LEDs rely on solution spin-coating. However, this technique struggles to meet the demands of commercial production for pixel patterning and large-area fabrication, necessitating the selection or development of suitable patterning processes [
18–
20]. Benefiting from the low-temperature synthesis method of perovskite, numerous existing patterning strategies apply to PeLEDs, including inkjet printing, laser direct writing, nanoimprinting, and vacuum deposition [
21,
22]. Each technique possesses distinct advantages, necessitating further research and optimization by relevant researchers to realize future commercial active-matrix PeLED displays. Third, and a critical factor constraining the commercial production and application of PeLEDs, is the biological and environmental compatibility of lead [
23,
24]. Heavy metal lead (Pb) causes irreversible harm to both humans and the environment. Particularly in lighting and display scenarios where human exposure is likely, Pb content must be strictly limited [
25]. To address lead toxicity concerns, researchers have explored several engineering strategies, including lead immobilization, encapsulation to reduce lead leakage, and recycling of lead-based devices [
26−
28]. These approaches have achieved promising progress in minimizing lead release caused by moisture exposure or decomposition during device operation, without significantly compromising device performance. However, despite these advancements, the development of lead-free PeLEDs is still regarded as the ultimate solution to eliminate lead toxicity concerns.
In this paper, we reviewed and summarized the remarkable achievements in PeLEDs, including core parameters such as the champion EQE, maximum brightness, and operational lifetime. We discuss key challenges hindering the commercialization of PeLEDs, including the lagging efficiency of blue PeLEDs, patterning strategies for active displays, and research progress on lead-free PeLEDs. Finally, we highlight three promising development directions for PeLEDs, which lay the foundation for future full-color, stable, and high-resolution perovskite-based display applications.
2 Performance table
All works in Table 1 derive from published journal articles before December 2025. The tables are basically classified by the EL peak, listing the performance of the champion device in different emission peaks, including EQE (without optical out-coupling), elements of the perovskite film, dimension, device structure, maximum brightness, current efficiency, operational lifetime, efficient area, FWHM, CIE coordinate, and publication date.
Table 1 and Fig. 1 summarize recent advances in high-performance PeLEDs spanning various wavelength bands from the visible to near-infrared regions. Researchers have systematically improved the quality of perovskite emissive layers through compositional engineering, crystallization control, defect passivation, and dimensional regulation of perovskites, which collectively contribute to the enhanced overall device performance [
29–
32]. Researchers also explored strategies such as interface engineering and light outcoupling optimization to further enhance device performance [
33–
35]. All achievements in Table 1 are realized by solution-processing methods. The highest efficiency records for red, green, and blue LEDs are 32.5%, 32.1%, and 26.4%, respectively [
36–
38]. Recently, the thermal evaporation technology has been regarded as a considerable method for the commercialization of PeLEDs in terms of large-area fabrication, reproducibility, and process compatibility [
39–
41]. Through optimization methods such as defect passivation and crystallization control, the highest efficiency records for thermally evaporated red, green, and blue PeLEDs reached 12.6%, 16.9%, and 10.4%, demonstrating the great potential [
42–
44].
Lead-free metal halide LEDs have garnered increasing attention owing to their reduced toxicity relative to lead-based PeLEDs [
24,
25]. However, their performance remains inferior to lead-based devices. Table 2 shows the performance progress of lead-free metal halide LEDs. Remarkably, the efficiency of blue metal halide LEDs has reached 7.9% using the vacuum evaporation method, which is more stable, lower cost, and also facilitates the way of device integration [
45]. Through further research and development, these lead-free metal halide LEDs could emerge as strong contenders for next-generation displays.
3 Challenges and outlook of PeLEDs
3.1 Challenges
3.1.1 Advanced blue PeLEDs
Three primary methods are typically employed to realize blue PeLEDs. The first involves engineering the halide composition to modulate the bandgap [
46,
47]. The second introduces large organic cations to induce dielectric confinement effects, thereby shifting the perovskite emission peak [
48,
49]. The third method controls the perovskite grain size to introduce strong quantum confinement effects [
50,
51]. These three methods are not mutually exclusive and are often used simultaneously to achieve better device performance [
38,
52]. While blue PeLEDs have promising properties such as bright blue light emission [
53,
54], there are still several issues before the commercialization application: 1) undesirable device efficiency caused by the chloride-induced low radiative recombination rate and poor crystallinity [
55–
57], 2) poor operational stability raised by severe ion migration, especially in mixed-halide perovskites [
9,
58]. These issues are generally considered to be significantly alleviated in pure bromide-based perovskites, which typically achieve blue emission through A-site and B-site engineering or by reducing dimensionality [
59,
60]. However, excessive ionic doping can inadvertently exacerbate lattice distortion, thereby compromising the PLQY. For example, partially substituting Cs
+ with smaller A-site cations like Rb
+ to achieve blue emission in CsPbBr
3 intensifies octahedral distortion [
61,
62]. Alternatively, complete replacement of Pb
2+ at the B-site—such as Eu
2+ in CsEuBr
3 or Ce
3+ in Cs
3CeI
6 can also achieve blue emission [
45,
63]. However, these lead-free systems often suffer from difficult carrier injection due to their wide bandgaps, poor carrier transport properties, and unfavorable energy level alignment with adjacent charge transport layers. Beyond lattice effects, the intrinsic nature of low-dimensional blue perovskites further complicates performance optimization. Their high surface-to-volume ratio inevitably leads to increased surface defect density, creating additional non-radiative recombination centers that severely limit device efficiency. This situation underscores the need for the development of multifunctional defect-passivating molecules and bandgap-matched structural designs to enable high-brightness, efficient, and large-scale blue perovskite LEDs.
3.1.2 Active-matrix display
An active-matrix display is a self-emissive device in which each pixel is independently addressed by a thin-film transistor (TFT) or complementary metal-oxide-semiconductor (CMOS) driving circuit. Compared with conventional LCDs, active-matrix displays possess the advantages of wider viewing angle, lower power consumption, and higher contrast [
64]. The transition from laboratory-scale prototypical LED devices to active-matrix displays is an essential step toward the commercialization of PeLEDs [
65]. To achieve active-matrix PeLED displays, it is critical to select an appropriate pattern strategy. In traditional semiconductor manufacturing processes, III-V LED display units and backplane circuits are typically connected via flip-chip bonding, which requires additional annealing processes and pressure application [
66]. Due to the low tolerance of perovskite for high-temperature annealing and pressure, promising approaches involve in situ fabrication of PeLEDs directly on TFT or CMOS backplanes, such as inkjet printing and thermal evaporation. Although numerous studies have successfully demonstrated monochrome PeLED active-matrix displays using spin-coating techniques, achieving subsequent full-color PeLED displays remains challenging [
12,
67–
69].
Various patterning techniques have been developed for realizing full-color PeLED active-matrix displays, each with distinct trade-offs in resolution, scalability, and material compatibility [
70]. Inkjet printing offers high material utilization and scalability to large areas, and recent advances have pushed its resolution to the sub-micrometer scale, while the inkjet-printed PeLEDs have achieved a surprisingly high peak efficiency of 21.73% [
18,
71]. Furthermore, inkjet printing, especially electrohydrodynamic jet printing, has enabled the realization of active-matrix Micro-PeLED displays [
72]. In inkjet printing, solvent evaporation significantly influences the crystallization process of perovskites, often necessitating the introduction of mechanisms like the Marangoni effect to suppress the coffee ring effect during droplet drying. Consequently, perovskite inks must exhibit excellent stability, facilitate film formation, and avoid compromising underlying circuits to achieve high-performance active-matrix displays [
17].
Thermal evaporation has long been the primary manufacturing route for OLEDs, featuring well-established processing standards and extensive industrial experience. Through optimization methods such as defect passivation and crystallization control, the highest efficiency records for thermally evaporated red, green, and blue PeLEDs reached 12.6%, 16.9%, and 10.4%, respectively [
42–
44]. Leveraging the advantages of full vacuum deposition, thermally evaporated PeLEDs (~30 μm) have also been successfully implemented in active-matrix displays driven by TFTs [
73,
74]. During thermal evaporation, PeLEDs undergo more complex interactions involving physical evaporation, gas-phase or surface chemical reactions, nucleation, and grain growth compared to OLEDs [
70]. Therefore, in addition to precise rate control, thermal evaporation of PeLEDs requires addressing the thermodynamic and kinetic control of the crystallization process to minimize non-radiative recombination within the perovskite [
42,
73,
75].
Photolithography offers high resolution (< 1 μm), making it ideal for active-matrix displays. However, its solvent-based processes necessitate complex overlay and alignment procedures when depositing full-color PeLEDs, increasing manufacturing complexity and reducing display panel yield rates [
76]. Transfer printing offers good resolution and material compatibility by physically transferring pre-formed perovskite films, but its scalability to large-area production remains challenging [
77]. Laser direct writing enables mask-free, high-resolution (~1 μm) patterning with minimal solvent exposure, offering good compatibility with perovskite materials, but its throughput is limited for mass production [
78]. Moreover, while existing techniques rely almost entirely on spin-coated perovskite precursors, laser direct writing struggles to meet the requirements for fabricating RGB pixels in full-color displays. At the current stage, no single technology can perfectly meet all the requirements for perovskite patterning in active-matrix displays. Combining the well-established OLED manufacturing processes with the excellent performance of patterned PeLEDs, inkjet printing and thermal evaporation are considered as promising techniques. Furthermore, the integration of different patterning methods may offer a viable path toward simultaneously achieving high resolution, large-area fabrication, and other essential requirements [
79].
3.1.3 Lead-free PeLEDs
Significant advances in the efficiency and stability of lead-based PeLEDs confirm their great potential as the next generation of environmentally friendly display technology. As environmental awareness grows, technologies relying on toxic materials are being phased out in favor of more sustainable alternatives [
25,
80,
81]. Consequently, exploring lead-free perovskite material systems and enhancing device performance are essential for commercialization. The focus of current research is tin, germanium, copper, and several other lead-free perovskites [
82–
84]. Of particular note are tin-based PeLEDs with more than 20% EQE, which are regarded as the most promising alternative to lead-based perovskites [
73]. Nevertheless, the challenges to be addressed if tin-based PeLEDs are to be rendered commercially viable include the poor stability and efficiency resulting from sensitivity to oxidation, uncontrollable crystallization processes, and energy level mismatches due to high conductivity [
24,
75]. The lifetime of lead-free PeLEDs is currently limited to tens of hours, yet device-level optimization, such as the use of multifunctional additives to modulate crystallization, offers viable pathways for improvement [
85].
While encouraging progress has been made with lead-free PeLEDs for red and near-infrared emission, there is still significant room for improvement in the green and blue spectral regions, as shown in Fig. 1. For most current lead-free LEDs, high efficiency is usually achieved at low current densities, while exhibiting significant efficiency roll-off that limits the device brightness [
24]. For lead-based PeLEDs, significant progress has been made in mitigating efficiency roll-off. Recent studies indicate that through crystallization engineering, ligand engineering, and the use of single-crystal perovskite emitting layers, lead-based devices can now maintain high efficiency even under conditions of high current density or high brightness [
86−
88]. Typically, lead-based PeLEDs operate at carrier densities below 10
15 cm
− 3 (corresponding to ∼600 mW·cm
−2), where Auger recombination remains relatively suppressed [
89]. In contrast, lead-free PeLEDs face more fundamental challenges due to their inherently higher carrier densities during operation. For example, in classical tin-based perovskites, the carrier density can exceed 10
18 cm
− 3 [
90]. This significantly higher carrier density makes lead-free devices more susceptible to Auger recombination under high excitation densities, leading to severe efficiency roll-off even at relatively low current densities [
24]. The strong Auger recombination has been confirmed as the primary cause of efficiency roll-off in tin-based perovskites [
91]. This fundamental limitation, combined with difficult electrical injection and high defect densities, results in both peak efficiency and achievable brightness remaining substantially lower than their lead-based counterparts. We propose that exploration of lead-free perovskite LEDs should begin with a step-by-step approach: partially substituting and alloying lead-based perovskites [
92,
93]. This strategy reduces Pb toxicity while minimizing impact on device performance, making them more suitable for practical display or lighting applications. This approach of transitioning from low-lead to lead-free systems will also provide valuable reference for exploring lead-free perovskite material systems and optimizing device structures.
3.2 Outlooks
3.2.1 Full-color display
The realization of full-color displays using PeLEDs is regarded as a transformative frontier in next-generation optoelectronics, driven by their exceptional color purity (FWHM < 20 nm), broadly tunable emission wavelengths (400−700 nm) via halide composition/alloying, and compatibility with solution-processed fabrication. Scalable patterning techniques and hybrid integration strategies could address challenges in pixel uniformity and color gamut consistency [
21]. Furthermore, photonic engineering may enhance light outcoupling and chromaticity control [
33]. Machine learning (ML) holds broad potential across the PeLEDs development pipeline, including additive screening, device structure optimization, performance simulation, and operational lifetime prediction [
94,
95]. By learning from experimental data, machine learning models can identify hidden correlations between composition, architecture, and device performance, guiding researchers toward more efficient and stable configurations [
96]. In the future, the convergence of ML-guided design and industrial-scale deposition methods will enable PeLED-based full-color displays to emerge as viable competitors to OLED and QLED technologies [
97,
98].
3.2.2 Micro/nanoscale display
Emerging display technologies, such as augmented reality, virtual reality, and mixed reality displays, demand miniaturization of LED units [
99,
100]. The high production costs and significant efficiency losses associated with decreasing dimensions of III–V semiconductor–based micro-LEDs (Micro-LEDs) have hindered their commercial application potential. In contrast, spin-coated PeLEDs exhibit smaller efficiency drops than micro-LEDs even as device feature sizes decrease from hundreds of micrometers to hundreds of nanometers [
101]. Furthermore, epitaxially grown micro-PeLEDs achieve a maximum EQE of 16.7% and high brightness of 4 × 10
5 cd·m
−2 at a 4-μm device size, highlighting the performance advantages and integration feasibility of micrometer- and nanometer-scale PeLEDs as next-generation light sources [
32].
3.2.3 Device stability
In practical applications, commercial settings demand long-term stable operation from displays, such as a basic
T50 exceeding 10,000 h at
L0 = 100 cd·m
−2 [
102]. Although significant progress has been made in the operational stability of PeLEDs in recent years, the performance gap among devices with different emission colors remains substantial, posing a critical bottleneck for full-color display applications. Currently, green PeLEDs demonstrate the best stability, with reported
T50 values at
L0 = 100 cd·m
−2 exceeding 50,000 h [
103], approaching the basic stability requirements for practical applications. Red PeLEDs have also shown continuous improvement in stability, with representative
T50 values now reaching over 100 h under the same measurement conditions [
104]. However, the stability of blue PeLEDs lags far behind. Despite ongoing research efforts in this area, the
T50 lifetimes of even the most advanced blue devices typically range from only a few minutes to a few hours. For example, early blue-emitting devices based on quantum confinement effects showed a
T50 of only about 250 s [
105]. More recent studies using defect passivation and bulky cation engineering for blue PeLEDs have improved the
T50 exceeding 5 h at
L0 = 1000 cd·m
−2 [
54]. Furthermore, the inconsistent lifetimes of the three RGB colors of PeLEDs can further impact the operational state of full-color displays. Therefore, the operational stability of PeLEDs remains a critical challenge in practical applications.
Key degradation pathways in perovskites, such as electric field-driven phase segregation, halide migration, and moisture-induced lattice collapse, can be effectively addressed through advanced material and device strategies [
106,
107]. These include low-dimensional perovskite engineering for improved thermodynamic stability, interface passivation using functional polymers or inorganic layers to suppress ion migration, and robust encapsulation technologies to block environmental stressors [
108,
109]. All-inorganic perovskites and lead-free alternatives show promise in addressing intrinsic instability and toxicity concerns. Concurrently, defect-tolerant synthesis techniques and optimized charge transport layers could minimize non-radiative recombination, extending operational lifetimes [
44]. Future advancements will likely focus on integrating machine learning-guided material discovery with scalable fabrication methods (such as inkjet printing, thermal deposition) to achieve both stability and performance metrics comparable to commercial OLEDs. Collaborative efforts across academia and industry are essential to standardize stability protocols (e.g., under continuous illumination, thermal cycling) and bridge the gap between laboratory breakthroughs and market-ready devices.