Progress and perspective on CsPbX₃ nanocrystals for light emitting diodes and solar cells

Lead halide perovskites have attracted great attention for optoelectronic applications in recent years,^1–10 mainly due to their superior optoelectronic properties such as high absorption coefficients,^11–13 relatively low carrier trap density,^14–15 long carrier diffusion lengths,^16–18 low exciton binding energies,¹⁹ and so on. Generally, lead halide perovskites possess a formula of ABX₃, where A is a monovalent cation [methylammonium (MA⁺), formamidinium (FA⁺), or cesium (Cs⁺)], B is a divalent metal cation (Pb²⁺ or Sn²⁺), and the X is a halogen anion (Cl⁻, Br⁻, I⁻, or mixed halides).^20,21 During the past several years, lead halide perovskites have not only pushed forward the development of cost-effective and high-performance photovoltaic technologies but also demonstrated great advances in applications such as light emitting diode (LEDs),^22–24 solar cells (SCs),^25,26 photodetectors,^27,28 and lasers.²⁹ Despite the great success, several problems remain to be resolved, including, in particular their long-term stability under ambient conditions. It has been found that the low intrinsic stability of perovskites originated from the organic groups (MA⁺, FA⁺), which is extremely sensitive to moisture, oxygen, and heat.^30,31 An effective way is to replace the organic group with inorganic cesium to produce all-inorganic perovskites. In 2015, Protesescu’s group pioneered the fabrication of cesium lead halide perovskites (CsPbX₃, X = Cl, Br, and I) by the hot-injection method, which exhibit high thermal stability and distinctive optical properties.³² Immediately after their report, many researchers became enthusiastic about the preparation, performance research, and application promotion of the all-inorganic perovskites.^33–43 In this perspective, we first discuss the structure and optoelectronic properties of the all-inorganic perovskite CsPbX₃ NCs [Fig. 1(a)]. Then, we mainly demonstrate recent advances in the fields of CsPbX₃ NC-based LEDs and perovskite solar cells (PSCs). Figure 1(b) shows the external quantum efficiencies (EQEs) and power conversion efficiencies (PCEs) of CsPbX₃ NC-based LEDs and SCs, respectively, which were demonstrated from 2015 to 2020 (the data were extracted from Tables I and II). Finally, we present a brief perspective including challenges and directions on all-inorganic perovskites.

First, to deeply understand the optoelectronic properties of CsPbX₃ NCs, it is necessary to be familiar with their crystal and electronic structures. CsPbX₃ crystals mainly have orthorhombic, tetragonal, and cubic phases.^20,32 The ideal phase structure of CsPbX₃ with superior performance is cubic. As shown in Fig. 2, for the cubic structure of CsPbX₃, Pb²⁺ coordinates with six X anions to form PbX₆⁴⁻ octahedron, and Cs⁺ is located at the center of the framework surrounded by eight PbX₆⁴⁻ octahedrons. Generally, the structural stability of ABX₃ perovskite can be largely determined by the ionic radii of A-site, B-site, and X-site ions that correspond to the Goldschmidt's tolerance factor (t), which can be expressed as⁴⁶ 

where R_A, R_B, and R_X are refer to ionic radii of A, B, and X ions, respectively. For stable ABX₃ perovskite structure, t should be in the range of 0.813–1.107, whereas 0.9 < t < 1 is generally considered a good fit for an ideal cubic perovskite structure.^47,48 Lowering or exceeding the upper limit of the t-factor may result in the non-perovskite phase of the perovskite. Additionally, all CsPbX₃ perovskites show direct bandgaps, and the electronic structures of CsPbX₃ perovskites with different halide compositions are found to show qualitatively common features, indicating their potential application in optoelectronics.^34,49

CsPbX₃ NCs possess a series of unique optical properties such as the following (Fig. 3):³² emission spectra can be tuned over the entire visible spectral region by adjusting their composition luminescence spectrum (410–700 nm) [Fig. 3(a)]; narrow emission linewidths of 12–42 nm [Fig. 3(a)], which are much narrower than that of traditional quantum dots (such as CdS and CdSe); wide color gamut (≈140%) [Fig. 3(b)]; and ultrahigh photoluminescence quantum yield (PLQY) that nearly reaches 100% [Fig. 3(c)].⁵⁰ Additionally, if the size of the CsPbX₃ NCs is down to the quantum confinement regime, these NCs exhibit a distinctive quantum confinement effect, which is closely related to the particle size. For example, when the size of CsPbBr₃ decreases from 11.8 to 3.8 nm, the emission wavelength changes from 520 to 460 nm due to the quantum confinement effect [Fig. 3(d)].³² Furthermore, the CsPbX₃ NCs show two-photon or three-photon absorption properties, showing great potential for applications in nonlinear photonics and optical devices.^51–52 Due to their ionic crystal characteristics and intrinsic vacancy defects, CsPbX₃ NCs have good ionic conductivity and carrier transport properties. The carrier mobility in CsPbBr₃ NCs exhibited one to three orders of magnitude higher than CdSe NCs, polycrystalline films, and single crystals of MAPbBr₃. Yettapu et al. have found that the carrier mobility and diffusion length of the CsPbX₃ NCs are up to 4500 cm²V⁻¹ s⁻¹ and 9.2 μm, respectively.⁵³ These amazing carrier transfer performances of CsPbX₃ NCs would benefit their promising applications in optoelectronic and photovoltaic devices.

CsPbX₃ perovskite LEDs have been intensively investigated by virtue of superior optical and electrical properties of CsPbX₃ NCs. Song et al. first reported CsPbX₃ NC-based LEDs (ITO/PEDOT:PSS/PVK/QDs/TPBi/LiF/Al) [Fig. 4(a)], which exhibit blue, green, and yellow emission and showed a modest EQE of 0.12% [Fig. 4(b)].⁴⁴ Taking advantage of the facile emission-color tunability, multicolored LEDs based on CsPbX₃ NCs with varying halide composition were realized [Figs. 4(c)–4(e)]. However, the non-optimized device exhibited poor efficiency and stability. Various factors must be taken into account when fabricating perovskite CsPbX₃ LED devices, such as optical properties, stability, and carrier injection of the devices. Thus, inspired by the pioneering work, tremendous efforts have been devoted to improving the device performance and stability of the CsPbX₃ NC-based LEDs, such as ligand modification, ion doping, and transport layer engineering.

High-quality CsPbX₃ NCs are crucial for fabricating high-performance LEDs. Surface passivation with long chain ligands is usually necessary to keep a high PLQY, which will hinder carrier injection of NC films. Therefore, replacing long ligands with relatively short ones is an effective way to improve charge carrier injection and maintain charge balance. Pan et al. exploited a relatively short ligand of di-dodecyl dimethyl ammonium bromide (DDAB) to passivate CsPbX₃ NCs and improve charge carrier transport.⁵⁴ CsPbX₃ NC-based LEDs displayed a maximum luminance of 330 cd m⁻² and a high EQE of 3.0%. Additionally, surface ligands of NCs have double-side effects on CsPbX₃ NC-based LEDs. More ligands are needed to provide sufficient surface passivation, but excessive ligands will cause poor electric conductivity. The low EQE of CsPbX₃ NC-based LEDs could be attributed to the excessive ligands that form an insulating layer. To solve this problem, Li et al. developed a new surface purification method to control the ligand density and balance the surface passivation and carrier injection, as shown in Fig. 5(a).⁵⁵ Finally, the CsPbX₃ LED device showed a maximum luminance of 15185 cd m⁻² and an EQE of 6.27% [Figs. 5(b) and 5(c)], which was 50-fold higher than that of the CsPbX₃ NC-based LEDs in their first report. Similarly, Chiba et al. reported a novel washing process using an ester solvent to remove excess ligands and achieved a highest EQE of 8.73% of CsPbBr₃ NC-based LEDs.⁵⁶ 

Doping CsPbX₃ with targeted ions could tune luminescence properties and electrical features, which would improve the efficiency and stability of CsPbX₃ NC-based LEDs. Transition metal element Mn is one of the most commonly used dopants for CsPbX₃ NCs. Zou’s group demonstrated an effective strategy through Mn²⁺-substitution to fundamentally stabilize lattices of CsPbX₃ NCs due to the enhancement of formation energy.⁵⁷ Benefiting from the Mn²⁺ doping, the EQE of CsPbBr₃ NC-based LEDs were promoted from 0.81% to 1.49% [Figs. 6(a) and 6(b)]. Lu’s group reported that Ag cathode was utilized to passivate surface defect states of CsPbI₃ NCs and simultaneously acted as dopant to reduce the electron injection barrier in CsPbI₃ LEDs.⁵⁸ The peak brightness of a CsPbI₃ NC LED device with an Ag cathode (1106 cd m⁻²) was more than double that of the ITO-based LED (455 cd m⁻²), and the current density of the Ag-based device was lower than that of the ITO-based device under the same voltage, as shown in Fig. 6(c), which demonstrated that higher charge injection efficiency was achieved benefiting from the reduction of the electron injection barrier. The average peak EQE of the LEDs increased from 7.3% to 11.2% upon replacing the ITO cathode with Ag [Fig. 6(d)]. Further on, they developed a method to improve the photo/electroluminescence (EL) efficiency and stability of CsPbI₃ NCs by using SrCl₂ as a co-precursor.⁵⁹ Sr²⁺ cations incorporate into the perovskite lattice to partially replace Pb²⁺ cations, resulting in slight lattice contraction and enhanced formation energy of perovskite NCs, which improves their stability. At the same time, Cl⁻ anions from SrCl₂ are able to efficiently passivate surface defect states of CsPbI₃ NCs, thus converting nonradiative trap states to radiative states, leading to the enhanced PLQY and prolonged PL lifetime. As shown in Fig. 6(e), the maximum values of luminance are 510 cd m⁻² and 1152 cd m⁻² for the CsPbI₃ and CsPbI₃-0.1 (the synthetic ratio of SrCl₂:PbI₂ equal to 0.1) based LEDs, respectively. LEDs using CsPbI₃-0.1 NCs as the emitting layer show a significantly enhanced device performance with a high EQE of 13.5% [Fig. 6(f)]. Furthermore, organic groups (FA⁺, MA⁺) were used to reduce the trapping of excitons and enhance the recombination efficiency of electrons and holes in CsPbX₃ NC films. Song et al. demonstrated that the LEDs based on CsPbBr₃ NCs exhibit a peak EQE of 11.6% and a current efficiency of 45.4 cd A⁻¹ after introducing a small amount of FA⁺ dopant.⁶⁰ 

To make highly efficient and stable perovskite-based LEDs, appropriate carrier transport layers are important. Imbalanced charge carrier injection would cause electrons’ or holes’ accumulation, which consequently results in non-radiative recombination. The effective way to solve this problem is carefully selecting and modifying the electron-transporting layer (ETL) and hole-transporting layer (HTL). Choosing suitable materials with energy level matching is necessary for selecting and modifying the ETL and HTL to obtain high efficiency of the device. Good electron transport capability and energy level match for electron injection are important to ETL, while good hole transport capability and energy level match for hole injection are important to HTL. Considering the inherent chemical instability of polymers in the transporting layer, selecting stable inorganic semiconductors for ETL and HTL is effective for improving the stability of CsPbX₃ devices. Chemical doping in the ETL or HTL is an effective technique to tune the energy levels and to balance the electrons and holes in the CsPbX₃ perovskite emissive layer. Besides, it is a feasible approach to introduce a charge transport interfacial layer between the transport and emitting layers. Zhang et al. introduced a perfluorinated ionomer (PFI) thin film sandwiched between the HTL and the CsPbBr₃ NCs layer, which facilitated the hole injection and prevented the NC emitters from charging.⁶¹ Wei et al. have presented an approach to control the perovskite compositional distribution and for efficient green-emitting CsPbBr₃ NC-based LEDs.⁶² They also demonstrated that the incorporation of an insulating layer of poly(methylmethacrylate) (PMMA) is critical to overcome the imbalanced charge carrier injection, which further increases the EQE from 17% to 20.3%. Additionally, inorganic charge transport layers possessed high resistance against moisture and oxygen and show better environmental stability and higher charge mobility compared with the traditional organic materials. In this regard, Shi et al. fabricated a stacking all-inorganic multilayer structure by using perovskite CsPbBr₃ NCs as the emissive layer and inorganic n-type MgZnO and p-type MgNiO as the carrier injectors to replace the organic carrier-transporting layers, as shown in Fig. 7(a).⁶³ Through energy band engineering of carrier injectors by Mg incorporation and their thickness optimization, CsPbBr₃ NC-based LEDs produced a maximum luminance of 3809 cd/m², luminous efficiency of 2.25 cd/A, and EQE of 2.39% [Fig. 7(b)]. Continuation of the above work, they demonstrated a strategy of solution-processed all-inorganic heterostructure that was proposed to overcome the emission efficiency and operation stability issues facing the challenges of perovskite LEDs.⁶⁴ Solution-processed n-ZnO nanoparticles and p-NiO are used as the carrier injectors to fabricate all-inorganic heterostructured CsPbBr₃ NC-based LEDs [Fig. 7(c)]. The device with bright green emission [Fig. 7(d)] produced a maximum luminance of 6093.2 cd/m², an EQE of 3.79%, and a current efficiency of 7.96 cd/A [Fig. 6(e)]. More importantly, the studied CsPbBr₃ LEDs possess good operation stability after a long test time in air ambient and can endure a high humidity (75%, 12 h) even without encapsulation [Fig. 7(f)]. Thus, this work points out the direction for high stability inorganic perovskite LEDs approaching the standards for practical use.

Table I shows a summary of the performances for CsPbX₃ NC-based LEDs. From the above discussion, we can conclude that although great progress has been made in the study of CsPbX₃ NC-based LEDs during the past five years, there are still many problems for practical applications. Their efficiency, cost, and stability need further improvement before practical applications.

In the last few years, organic–inorganic hybrid PSCs have had unprecedentedly progress from an unstable 3.8% PCE to over 25%.⁷⁸ Despite their high performance, these PSCs are thermally unstable and moist owing to the volatility of their organic components, which greatly hinder their further commercial applications.⁷⁹ Replacing the organic cation with inorganic Cs⁺ has been found to be effective and improved the stability. In 2015, Kulbak et al. first fabricated CsPbBr₃ PSCs with a PCE of 5.95% and an open-circuit voltage (V_OC) of 1.28 V, which showed excellent performance equally well with the MAPbBr₃ PSCs.⁸⁰ The CsPbBr₃ layers were annealed at 250 °C and the cells were fabricated in an ambient atmosphere, implied a fairly good stability. Though the PCE in this work is not high, this pioneering research showed that an all-inorganic version of the perovskite material works equally well as the organic one, in particular, generating the high open-circuit voltages that are an important feature of these cells. Since then, many researches have focused on inorganic CsPbX₃ PSCs.^81–85 Achieving high PCE, the optical bandgap of the absorption layer is of vital importance. The bandgap of CsPbX₃ is decreased from 3.0 to 1.73 eV with the halide ions from Cl, Br, to I. For CsPbCl₃, the optical bandgap is above 3.0 eV, which makes it unsuitable for solar cell applications. CsPbI₃ NCs are often better characterized as the gamma (γ) perovskite phase.⁸⁶ However, black cubic (α) CsPbI₃ with a suitable bandgap of 1.73 eV may be the best candidate for CsPbX₃ PSCs. Despite the fact that CsPbI₃ PSCs can already achieve PCE as high as 19%,⁸⁷ CsPbI₃ will immediately transform into the yellow orthorhombic (δ) phase with a wide bandgap of 2.82 eV at a temperature below 315 °C, which basically does not absorb much sunlight and impedes their applications. Additionally, many physical properties differ between nanometer-sized and bulk crystalline materials of the same chemical compound. CsPbX₃ NCs exhibit improved room temperature cubic-phase stability and attractive optical properties.⁴⁵ Interestingly, CsPbX₃ NCs provide a reduced crystallite grain size, which stabilizes the desired cubic phase without suffering from a bandgap increase upon anion mixing.⁸⁸ Thus, an effective way to stabilize PSCs is to form CsPbX₃ NC-based PSCs. In this prospect, we mainly introduce the research progress of CsPbX₃ NC-based PSCs.

The modification of CsPbX₃ NCs is essential for high-quality CsPbX₃ NC-based films’ formation that results in more efficient charge transport. Many research studies have been devoted for NC modification, such as FAI (formamidinium iodide) coating, micrometer-sized graphene (μ-graphene, μGR) cross-linking, and GeI₂ additives. In 2016, Swarnkar and co-workers developed an approach to purify CsPbI₃ NCs using methyl acetate (MeOAc) as an antisolvent, and these CsPbI₃ NCs remained stable in the desired cubic phase (E_g = 1.73 eV) for months of storage in an ambient.⁴⁵ The champion solar cell employing 9 nm α-CsPbI₃ NCs as a light absorber showed a respectable PCE of 10.77% and an open-circuit voltage (V_OC) of 1.23 V for a 0.10 cm² cell made and tested completely in ambient conditions (relative humidity ∼15% to 25%), as shown in Figs. 8(a) and 8(b). The low charge carrier transport efficiency of NC-based films hindered the efficiencies of the PSCs. Later on, the same group explored an A-site cation halide salt (AX) treatment (where A refers to either Cs, MA, or FA cations) for CsPbI₃ NCs.⁸⁹ The AX treatment was presented to double the film mobility, enabling increased photocurrent, and lead to a record certified PCE of 13.43%. In addition, phase change induced by agglomeration of NCs will result in low carrier transport efficiency. To address this problem, Wang et al. used high mobility (800 cm² V⁻¹ s⁻¹) micrometer-sized μ-graphene (μGR) sheets to cross-link the CsPbI₃ NCs for optimal electronic conductivity and at the meantime to keep NCs from agglomeration to achieve long-term stability.⁹⁰ The μGR provides effective carrier transport channel resulting in increased PCE to as high as 11.40%, nearly 12% improvement compared to the 10.17% efficiency achieved by the control device. In 2019, Liu and co-workers have showed that CsPbI₃ NCs with near-unity PLQY and long chemical stability can be obtained by using PbI₂/GeI₂ in combination with trioctylphosphine (TOP) ligands.⁹¹ CsPbI₃ NC-based PSCs (Au/Spiro-OMeTAD/CsPbI₃ NCs/TiO₂/FTO) [Fig. 8(c)] delivered a PCE of 12.15% [Fig. 8(d)] and retained 85% of their peak performance after storage over 90 days [Fig. 8(e)]. Also, proper ligand amount is necessary for the excellent device performance. Low ligand amount will induce serious trouble in film formation, and high ligand amount will lead to poor electric transport ability. More recently, Liu et al. have showed an effective MeOAc solvent treatment method to obtain a high-quality film with good charge transportation by controlling the amount of the ligands around the initial CsPbBrI₂ NCs in the film.⁹² Through this method, the trap state density decreased about four folds and the carrier mobility increased nearly 15 folds. They demonstrated that controlling the ligand amount around the NCs at 1.01 wt. %, the PSC exhibits a PCE as high as 12.2%.

Improving charge transport within the NCs and carrier extraction at the interfaces with the charge transport layers are crucial for the performance of PSCs. Yuan et al. have fabricated solution-processed FTO/TiO₂/CsPbI₃ NCs/hole-transporting material (HTM)/MoO₃/Ag PSCs at room temperature under ambient conditions, employing a series of dopant-free polymeric hole-transporting materials (HTMs).⁹³ CsPbI₃ NC-based PSCs showed efficient charge extraction at NC/polymer interfaces and avoid device instability by using the polymeric hole conductor PTB7, achieving a remarkable PCE up to 12.55% and an extremely low energy loss of 0.45 eV, as shown in Fig. 9(a). The large energy losses (E_loss) have become a major hindrance for the ultimate efficiency of CsPbX₃ PSCs. Zeng et al. have reported an effective and reproducible method of modifying the interface between a CsPbI₂Br absorber and poly(3-hexylthiophene) (P3HT) hole-acceptor to minimize the E_loss.⁹⁴ They demonstrated that electron–hole recombination could be reduced by depositing a thin P3HT layer on the top of CsPbI₂Br NC films [Fig. 9(b)], due to the electronic passivation of surface defect states. The CsPbI₂Br NC-based PSCs with annealing P3HT showed a V_OC of up to 1.32 V and E_loss of down to 0.5 eV, with a highest PCE of 12.02% [Fig. 9(c)]. In addition, the optimal devices show excellent stability retaining around 90% of their initial PCE after aging for near 1000 h [Fig. 9(d)].

Notably, CsPbX₃ NCs are appropriate for use as an interface engineering layer in PSCs due to their superior stability and tunable bandgap position. Liu et al. demonstrated that the efficiency and stability of PSCs can be enhanced by introducing stable α-CsPbI₃ NCs as an interface layer between the MA_0.17FA_0.83Pb(I_0.83Br_0.17)₃ (marked as FAMAPbI₃) film and the HTM layer.⁹⁵ By using a mixed-type perovskite FAMAPbI₃ as the light-absorption layer to further fulfill the valence band position (VBP) matching requirement between FAMAPbI₃ and CsPbI₃ NCs layers, the PCE of SC (FTO/c-TiO₂/m-TiO₂/FAMAPbI₃ film/CsPbI₃ NCs/Spiro-OMeTAD/Au) was increased from 15.17% to 18.56% due to the improvement of hole transmission efficiency from perovskite to HTM layer. Moreover, PSCs with CsPbI₃ NCs maintained 82% of their initial performance for 30 days at over 30% RH without any encapsulation, suggesting that the stability of the PSCs was also markedly improved due to the high stability of the CsPbI₃ NCs.

Up to now, the certified record PCE of CsPbX₃ NC-based PSCs (ITO/SnO₂/Cs_1−xFA_xPbI₃ NCs/Spiro-OMeTAD/Au) is 16.6% with negligible hysteresis, which reported by Hao et al., as shown in Figs. 10(a) and 10(b).⁹⁶ They developed an effective oleic acid (OA) ligand-assisted cation-exchange strategy to facilitate Cs_1−xFA_xPbI₃ NC-based PSCs with remarkable ambient stability. In an OA-rich environment, the cross-exchange of cations is facilitated, enabling the rapid formation of Cs_1−xFA_xPbI₃ NCs with reduced defect density. They further demonstrate that NC-based devices exhibit substantially enhanced photostability compared with thin film-based devices [Fig. 10(c)], due to the suppressed phase segregation. The mixed Cs_1−xFA_xPbI₃ NC system offers a pathway toward next-generation photovoltaics and optoelectronics with improved stability and flexibility. Finally, the advancement in the progress of CsPbX₃ NC-based PSCs in the past few years is shown in Table II. We believe that CsPbX₃ NC-based SCs may quickly emerge as strong contenders in the photovoltaic area.

In this perspective, we summarize the recent advances in CsPbX₃ NC-based LEDs and SCs. CsPbX₃ NCs not only have excellent optoelectronic properties but also have better stability, which makes a tremendous improvement in the performance of CsPbX₃ NC-based devices within the past few years. Despite great progress, there is still plenty of room for further improvement of CsPbX₃ NCs.

First, the prominent issue of CsPbX₃ NCs is their structural instability against moisture, oxygen, thermal, and photo-disturbances. Therefore, ligands, surface modification, or inert shells of CsPbX₃ NCs are still needed to further improve their stability. The main obstacle to the widespread applications of CsPbX₃ NCs is their inherent instability caused by ionic characteristics. Grain boundaries are abundant in NCs, thus providing ample pathways for ionic migration. Considering the success in quantum dot light emitting diodes (QLEDs), we think that it would be a good idea to coat CsPbX₃ NCs with well-designed semiconductor shells to overcome the low stability. The semiconductor shell can not only confine the excited carriers but also prevent the degradation during the operation process. Owing to the ionic nature of CsPbX₃ NCs, a CsPbX₃/semiconductor core/shell structure is challenging. A new preparation strategy for this core/shell structure is still anticipated.

Second, the EQEs of perovskite LEDs are over 20% for green, red, and near-infrared (NIR) regions with different device optimization strategies. Despite the great success of green or NIR perovskite LEDs, blue emitting devices have become the key shortcoming for wide-color-gamut display fields. The blue wavelength emission needs the most work. Using mixed chloride and bromide is an effective way to realize blue emission. Blue emissive from sky-blue (490 nm) to deep blue or even ultraviolet (410 nm) can be achieved by increasing the chloride content. However, the spectrum is unstable in the mixed halide LED system, which attributes to phase separation induced by the halide migration under an electric field. Thus, the color stability has been hit. Additionally, a larger charge injection barrier for blue perovskite LEDs than visible perovskite LEDs due to the deep valence band leads to poor charge injection efficiency and degradation of the overall device performance. Color instability and poor charge injection restrain the development of high-efficiency blue perovskite LEDs, which are the obstacles to the commercial applications. Particularly for CsPbX₃ NC-based blue LEDs, the highest EQE is 12.3% as reported by Dong et al.,⁷⁷ which is still far behind that of commercial organic LEDs and QLEDs. Hence, the development of blue LEDs is still in its infancy, and efforts should be made for CsPbX₃ NC-based blue LEDs. Although it is still some distance away, we believe that blue perovskite LEDs will be in commercial applications in the future.

Third, the necessary toxic lead element raises a critical concern for future commercial development. To address the toxicity issue, recent research effort has been devoted to develop lead-free halide perovskite (LFHP) NCs, such as Sn, Sb, and Bi based LFHP NCs. Up to now, for LFHP solar cells, the highest PCE is 13.24% reported by Nishimura et al., which is still a certain distance with lead-based perovskite.¹⁰⁸ Additionally, it remains challenging for LFHP NCs due to poor stability when dealing with heat, light, oxygen, and humidity. Looking into the future, many strategies may be pursued for improving the stability of LFHPs NCs, such as core/shell structure, ligand modification, air-stable metal dopant, and so on. Although the properties of LFHP NCs may be inferior to CsPbX₃ NCs, optical and device performance improvement of LFHP NCs is an important research direction in the future.

Although challenges remain, we believe that new and interesting physical and chemical properties will be found for all-inorganic perovskite NCs, and new devices will be developed for real life applications in the near future.

This research was funded by the National Natural Science Foundation of China (NNSFC) (Nos. 51872161, 51902179, and 11904198) and the Natural Science Foundation of Shandong Province (Nos. ZR2017ZB0316 and ZR2019BF019).

There are no conflicts to declare.

The data that support the findings of this study are available within the article.