Research related to perovskites can be traced back to 1970s,^1–3 but systematic research was lacking due to technology limitations in that period. In 2009, Kojima et al.⁴ first added organic–inorganic hybrid perovskites as semiconductor materials in dye-sensitized solar cells, achieving a power conversion efficiency (PCE) of 3.8%. Since that breakthrough, the development of perovskites with large absorption coefficient, low defect state density, long carrier diffusion length, and bipolar carrier transport property has made them uniquely suitable for photovoltaic applications.^4–16 Currently, the PCE of single-junction pure perovskite-based solar cells has reached 25.5% for small-area devices and 24.2% for large area over 1cm2.^15,17 According to the Shockley–Queisser limit, a type of high-quality conversion material in solar cells is also efficient luminescent materials in light-emitting devices such as LEDs and lasers.^18–20 In 2004, the first evidence of optical gain in lead halide perovskites was reported, which is amplified spontaneous emission (ASE) from microcrystalline films of CsPbCl3 recrystallized from the amorphous phase.^21,22 In 2014, ASE and lasing were realized from MAPbX3 polycrystalline thin films at room temperature. Ultralow threshold could benefit from the excellent optical absorption of MAPbX3 with a coefficient greater than 2×104cm−1.^23–27

In addition, the research about micro/nanolasers based on perovskite with high coherence, low threshold, and high-quality factor has increased rapidly. The advances in lasing performance mainly benefit from the excellent optical properties such as high photoluminescence quantum yield (PLQY), narrow linewidth, large absorption coefficient, and widely tuned band.^28–32 In addition, the shape and size of perovskite could be flexibly adjusted, which can affect their physical and chemical properties and the performance of optoelectrical devices.^33–36 Hence, various synthesis strategies about the fabrication, control of morphologies, and sizes of perovskite nanocrystals (NCs) have been developed. The size can be adjusted from several nanometers to microns, and the morphologies can be controlled as zero-dimension (0D) quantum dots (QDs), one-dimensional (1D) nanorods (NRs) and nanowires (NWs), two-dimensional (2D) nanoplates (NPs) and nanosheets (NSs), and three-dimensional (3D) nanocubes and microspheres (MSs).^37–42 In this review, we discuss and summarize the recent developments in lead halide perovskite materials and perovskite-based lasers. In particular, the synthetic strategies of the perovskite containing solution process and vapor evaporation method are reviewed. Moreover, the morphology control of perovskite with various dimensions for the natural resonant cavities of lasing is discussed. Based on the perovskite QDs, NWs, NRs, NPs, and MSs, the single perovskite nano/microlaser and laser array are reviewed, and the dependence of laser performance on structure morphology is discussed. Finally, we present a summary and the perspectives of future research in the perovskite-based laser.

As a kind of chemical material with an ABX3-type structure ^[Fig. 1(a)], the crystal structure of perovskite is the same as calcium titanate (CaTiO3).⁴³ “A” could be an organic molecular group such as methylamino (MA) or an inorganic element such as cesium (Cs); “B” is generally a metal ion, such as lead (Pb), tin (Sn), and bismuth (Bi); “X” refers to a halide ion containing Cl, Br, and I. The tolerance factor (t) is generally used to evaluate the structural formability and stability, calculated as t=(rA+rX)/2(rB+rX), where rA, rB, and rX are ionic radii of A, B, and X sites, respectively. Li et al.⁴⁵ introduced the “octahedral factor” (μ) to investigate the regularities of formability for cubic perovskite ABX3, which was defined as μ=rB/rX and suggested the formation of the halide perovskites under the conditions of 0.813<t<1.107 and 0.377<μ<0.895. Then, Sun et al. introduced (t+μ)η to evaluate the thermodynamic stability of hailde perovskite, where η is the atomic packing fraction in a crystal structure. By calculating decomposition energies of 138 perovskite compounds ^[Fig. 1(b)], they demonstrated better accuracy of (t+μ)η than the evaluation of t and μ alone.^44,46

Based on the ABX3 halide perovskites, the morphologies can be controlled with different dimensional nanostructures, such as 0D QDs, 1D NWs, 2D NPs, and 3D MSs.^{37,40,47–65} Since different nanostructures will result in variable structure–property relationships at the nanoscale level, various strategies have been reported for controlling the form and size of the perovskite NCs, including changing the reaction temperature, the reaction time, and the ligand combinations during synthesis.^37–42 In addition, capping ligands with different structures and lengths can also affect the nucleation and growth rate, hence the structure of perovskite NCs can be adjusted owing to their anchoring and steric effects.^47–49

As is well known, traditional semiconductor QDs have a quantum size effect. In the case of perovskite QDs, the bandgap can also be tuned via halide component regulation.^38–42,51 Zhang et al.²⁵ developed a ligand-assisted reprecipitation (LARP) method to fabricate MAPbBr3 QDs at room temperature. Their PLQYs were up to ∼70% ^[Fig. 2(a)]. By mixing PbX2 salts into the precursors, the authors tuned emission in the range of 407 to 734 nm ^{[Figs. 2(c) and 2(d)]}.²⁵ In the case of all-inorganic ones, the emission spectra of CsPbX3 QDs fabricated by the high temperature method could be tunable over 410 to 700 nm with narrow half maximum of 12 to 42 nm and radiative lifetime of 1 to 29 ns ^{[Figs. 2(e) and 2(f)]}.⁴² Subsequently, they proposed an anion-exchange process to tune the emission of colloidal CsPbX3 QDs via postsynthetic reactions with different compounds ^{[Figs. 2(g) and 2(h)]}.⁴¹ Besides the hot-injection technique, the room-temperature synthesis for perovskite QDs was also studied.³⁹ In 2016, Zeng and coworkers developed a room-temperature method to fabricate CsPbX3 QDs via supersaturated recrystallization. In this process, the crystallization process occurred in the transform of Cs+, Pb2+, and X− ions from soluble to insoluble solvents in the absence of inert gas within a few seconds, as shown in Figs. 2(i) and 2(j).³⁹ Although crystallized at room temperature, these CsPbX3 QDs held superior optical properties with PLQYs above 70%, and PLs can remain at ∼90% after aging 30 days in the air. Except for being regulated by changing composition, the bandgap of perovskite QDs also can be tuned by the size regulation of QDs. Chen et al.⁶⁶ fabricated CsPbBr3 QDs with an average diameter from 7.1 to 12.3 nm by modifying the temperature, and the corresponding PL emission peaks could be tuned from 493 to 531 nm. Fang et al.⁶⁷ synthesized MAPbBr3 QDs with tunable average diameter from 2.82 to 5.29 nm by varying the additive amount of surfactant, and the corresponding PL emission peaks could shift from 436 to 520 nm due to the quantum confinement effect. Most recently, there have been much more researches about perovskite QDs by composition engineering for wider optoelectronic applications.^68–70

Perovskite 1D NWs and NRs are more applicable in the field of optoelectronic applications due to their special anisotropic structures. In the growth of 1D perovskite structures, reaction temperature, reaction time, and organic ligands are critical factors for crystallization.^71–74 Deng et al.⁷¹ first fabricated MAPbI3 NWs via the one-step solution method. In this process, the precursor solution containing PbI2 and CH3NH3I was dropped onto a substrate and then heated at different temperatures. Finally, uniform MAPbI3 NWs were obtained after heating at 80°C for 10 min. In 2017, they fabricated Csx(MA)1−xPbI3 NWs through a two-step solution method.⁷² As shown in Fig. 3(a), PbI2 powder was dissolved in water at 75°C initially, then PbI2 separated out when the solution cooled down to room temperature. With the addition of CsI and MAI, perovskite NWs could be formed after shaking for a few seconds. The length and diameter of obtained perovskite NWs could reach 10μm and several hundred nanometers. Moreover, the amount of perovskite NWs was related to the concentration of PbI2 separated out from aqueous solution.⁷² Zhu et al. developed a direct conversion of MAPbI3 film into NWs through a recrystallization process ^[Fig. 3(b)]. The first step was the formation of perovskite film from a mixture of PbCl2 and CH3NH3I.⁷³ Then, a mixture solution containing DMF and isopropyl was dropped onto the as-grown perovskite film. Along with the evaporation of the solvent, NWs could be formed ^[Fig. 3(b)].⁷³ Furthermore, they found the content of DMF in isopropyl, and the rotation speed could affect the sizes of prepared MAPbI3 NWs.

As for all-inorganic perovskite, in 2015, Yang and coworkers used a solution method to synthesize single-crystalline CsPbX3 NWs first. The reaction temperature was set as 150°C to 250°C.⁵¹ They found that the reaction time was critical to the growth of NWs. As shown in Fig. 3(c), the SEM of prepared CsPbBr3 with different reaction times showed perovskite nanocubes formed initially, then NS and NW formed at 90 min ^[Fig. 3(c)].⁵¹ In the formation of Cs-based perovskite NWs, surface ligands could affect the width and size. Imran et al.⁵⁶ tuned the width of CsPbX3 NWs from 10 to 20 nm by regulating the ratio of octylamine to oleylamine and varying the reaction time. They found that the width of NWs can be decreased below ∼5nm by introducing carboxylic acids with short aliphatic chains. Correspondingly, the emission spectra of CsPbBr3 NWs could be tuned from 524 to 473 nm ^[Fig. 3(d)].⁵⁶ Amgar et al.⁵⁸ found that various hydrohalic acids (HX, X = Cl, Br, and I) affect the length of CsPbBr3 NWs efficiently. With the increasing amount of HX, the length of NWs would be shortened ^[Fig. 3(e)]. Using this method, the emission of the CsPbBr3 NWs could be tunable in the range of 423 to 505 nm ^[Fig. 3(f)].⁵⁸CsPbBr3 NWs/NRs can also be synthesized by a low-temperature method. Dong’s group⁷⁴ fabricated CsPbBr3 perovskite NRs with controllable size in a polymer matrix. Then, Liu et al. fabricated single-crystalline CsPbBr3 NWs without inert gas at room temperature. By increasing the reaction time, the length of NWs could be increased from nanometers to micrometers, and the diameter could be tuned from 2.5 to 32.0 nm. Moreover, using this method, the emission spectra of CsPbX3 NWs could be tuned from 434 to 681 nm.⁵⁷

Besides the above-mentioned solution-process, plenty of works have been reported on synthesizing perovskite NWs and NRs by vapor-phase growth.⁴⁹ More than ever, the vapor-phase process can control the morphology and crystalline phase of perovskite NCs efficiently. It has been demonstrated that the growth temperature and the substrates are critical for the orientation of perovskite NWs in vapor-growth. Xing et al.⁴⁹ first used the vapor-phase technique to fabricate perovskite NWs. First, PbI2 NWs were deposited on SiO2 substrates by the chemical vapor deposition (CVD) method ^[Fig. 4(a)].⁴⁹ Consequently, PbI2 was converted into MAPbX3 after the reaction with MAX through a CVD process. As shown in Fig. 4(b), the prepared MAPbI3 wires had a length about tens of micrometers and a diameter of ∼500nm. In Figs. 4(c) and 4(d), they indicated that MAPbI3 NW grows along the [100] direction.⁴⁹ Due to the thermal decomposition of organic hybrid perovskite occurring easily at high temperatures, direct vapor–phase growth of hybrid perovskites is more challenging. However, the vapor–phase technique is an attractive method for all-inorganic perovskites, which have better thermostability. Zhou et al.⁷⁵ prepared CsPbX3 NRs with high crystallization quality and regular triangular morphology through a vapor deposition method ^[Fig. 4(e)]. As shown in the SEM image ^[Fig. 4(f)], prepared CsPbX3 NRs had a triangular cross section, smooth surfaces, and a length of 2 to 20μm. They demonstrated that the reaction temperature was critical for the control of perovskite NCs during the growth of triangular CsPbBr3 NRs. Moreover, the emission of these as-grown CsPbX3 NRs also can be tuned from 415 to 673 nm by halide component regulation ^[Fig. 4(g)].⁷⁵

In addition, it has been confirmed that the substrate can affect the grain orientation growth of perovskite NWs.⁷⁶ Chen et al.⁶² fabricated CsPbX3 wires on mica by the CVD method. During the growth of CsPbBr3 NWs, heteroepitaxial matching occurred in the interface between CsPbBr3 NCs and mica substrate. Then, the formation of NWs was caused by the asymmetric lattice mismatch with the mica substrate. As shown in Figs. 4(h) and 4(i), the obtained CsPbBr3 wires were well-aligned, surface-bound, and formed a network with a length about tens of μm and width of ∼1μm, respectively.⁶² Moreover, various nanostructures could be formed by controlling the deposition time, such as single NWs, Y-shaped branches, and interconnected NW or MW networks.⁶²

The unique and excellent properties of 2D structured perovskite such as NSs, NPs, and microdisks (MDs) make them promising for potential optoelectronic devices.⁷⁷ Sichert et al.⁶¹ synthesized MAPbBr3 NPs and investigated the quantum size effect of NPs via the solution method. They found that the thickness of MAPbBr3 NPs was reduced with the increase of the content of OA ^{[Figs. 5(a) and 5(b)]}. Correspondingly, the PL emission was tuned from the green to violet region ^[Fig. 5(c)].⁶¹ Qin et al. prepared MAPbI3 NPs via a two-step solution method. First, PbI2/DMF solvent was spin-coated onto a substrate to form PbI2 thin films. Then, the formed PbI2 thin film was immersed into MAI solution, in which MAPbI3 single NCs were formed.⁸¹

CsPbBr3 NPs were prepared by Bekenstein et al.³⁷ through a hot-injection method. They demonstrated that the reaction temperature is critical for the shape and thickness of CsPbBr3 NPs. As the temperature decreased from 150°C to 130°C, the shape of CsPbBr3 NCs evolved from nanocubes to NPs. Correspondingly, the PL emission was shifted from 512 to 405 nm ^{[Figs. 5(d) and 5(e)]}.³⁷ When the temperature decreased to 90°C and 100°C, the thin CsPbBr3 NPs were obtained with lengths of about 200 to 300 nm ^[Fig. 5(d)].³⁷ Except for the reaction temperature, surface ligands also affect the formation of CsPbX3 NPs, which was demonstrated by Pan et al. in 2016. During the growth of NPs, CsPbX3 NPs were obtained at a relatively lower reaction temperature (120°C to 140°C). They obtained thinner CsPbX3 NPs with shorter chain amines.⁸² In addition, the reaction time was also found to be critical for the formation of perovskite NPs.^78,79 By adding the PbBr2 concentration and increasing the reaction time above 1 h (135°C), a CsPb2Br5 microplate (MP) with a micrometer order size and regular end faces could be obtained ^[Fig. 5(f)].⁷⁸ In 2018, Li et al. demonstrated that 2D CsPbX3 NPs and NSs can be obtained by varying the reaction time ^[Fig. 5(g)]. The thickness can be controlled in the range of 3 to 6 nm and the width in the range of 0.1 to 1μm.⁷⁹ Huang et al. reported a method for spontaneous crystallization of perovskite NCs in nonpolar organic solvent by mixing precursor ligand complexes without any heat treatment. By varying the ratio of monovalent to Pb2+ cation–ligand complexes, the shape of the NCs can be controlled from 3D nanocubes to 2D nanoplatelets.⁸³

Similar to the NWs, perovskite NPs can also be formed by the vapor synthesis method. Xiong and coworkers⁸⁰ reported the CVD growth of MAPbI3 NPs. These NPs exhibited triangular or hexagonal platelet shapes, with thickness of 10 to 300 nm and lateral dimensions of 5 to 30μm ^{[Figs. 5(h) and 5(i)]}. PbX2 platelets were first grown on mica via van der Waals epitaxy and then converted to MAPbX3 NPs with the existence of MAX. In 2016, Bao and coworkers developed a combined method containing a solution process and a vapor-phase conversion process to prepare MAPbI3 NSs. First, PbI2 flakes were dropped on a silica substrate and then heated. In this process, the temperature plays a crucial role in the nucleation and growth of 2D PbI2 NSs, since the amount of nucleation sites is controlled by temperature. Subsequently, MAPbI3 NSs were formed after the conversion reaction with MAI.⁸⁴ During the vapor-phase growth, the growth pressure and temperature both could affect the formation of perovskite NCs. Liu et al.⁶⁴ fabricated 2D MAPbBr3 platelets (001) via the CVD method. As shown in Fig. 5(j), the square-shaped platelets could not form, as the growth pressure and temperature were low. By increasing the pressure, 2D platelets and 3D spheres could be observed. The average thickness of MAPbBr3 platelets increased from 29 to 73 nm, and the lateral size increased from 6 to 10μm with the pressure increasing from 140 to 200 Torr.⁶⁴

As for all-inorganic 2D perovskite NCs, Zeng and coworkers⁸⁵ synthesized ultrathin CsPbBr3 NPs (thickness ∼148.8nm) on a mica substrate by van der Waals epitaxy through heating the PbBr2 and CsBr mixture. Zheng et al.⁸⁶ synthesized 2D CsPbI3 perovskite NSs with high quality, controllable morphology, and ultrathin thickness (∼6.0nm) via a space-confined vapor-phase epitaxial growth. In 2020, Yang and coworkers developed a facile method to pattern CsPbX3 plate arrays with crystal size (200 nm to 1μm) and spacing (2 to 20μm). These plate arrays were confined by prepatterned hydrophobic/hydrophilic surfaces.⁶⁵ The method can evade the restriction of lattice matching between perovskite and substrates, enabling a large-area growth of 2D perovskite NCs with excellent crystalline quality.⁸⁵

A metasurface is a type of 2D optical element composed of units with subwavelength scale size, producing resonant coupling between electric and magnetic components of the incident electromagnetic fields.^87–91 Several functionalities were demonstrated on all-dielectric metasurfaces, such as optical encoding, optical wavefront molding, polarization beam splitter, and enhanced PL.^92–94 Perovskite-based metasurfaces demonstrated potential for nonlinear absorption and optical encoding.⁹⁵ Metasurfce structures can be realized by nanopatterning thin film. Many conventional nanofabrication techniques have been used for the fabrication of perovskite metasurfaces, such as nanoimprinting, electron beam lithography (EBL), focused ion beam milling (FIB), and inductively coupled plasma etching (ICP).^{90,94,96–98}

Gholipour et al.⁹⁷ first used the FIB technique to fabricate MAPbI3 metasurfaces (thickness ∼200nm), which consisted of nanogratings and nanoslit metamolecules. Moreover, they demonstrated that the emission and quality factor of the reflection resonances can be tuned by varying the grating period.⁹⁷ Makarov et al.⁹⁸ developed nanoimprinting technology for patterning CsαFAβMAγPb(IxBry)3 metasurfaces, enabling them to enhance their linear and nonlinear PL ^{[Figs. 6(a)–6(d)]}. After the spin-coating of the perovskite film with thickness of ∼200nm, nanoimprinting with nanopillar and nanostripe molds was performed on perovskite thin film to form metasurfaces. They demonstrated that these metasurfaces can enhance linear PL eight times and nonlinear PL 70 times.⁹⁸ Jeong et al.⁹⁴ presented a polymer-assisted nanoimprinting method for fabricating large-area CsPbX3 nanopatterns. As shown in Fig. 6(e), during their nanoimprinting process, a precursor solution was spin-coated on a substrate initially, and then the nanoimprinting mold was pressed on the precursor film with thermal treatment subsequently. Thus, CsPbX3 was crystallized within the confines of molds ^[Fig. 6(f)]. This method could be easily extended to large-area perovskite patterns on different substrates.⁹⁴ In addition, Fan et al.⁹⁰ used the EBL flowed ICP technique to prepare near-infrared MAPbBr3 perovskite metasurfaces ^[Fig. 6(g)]. Based on these metasurfaces, many types of nonlinear processes and enhanced PL could be observed ^[Fig. 6(h)].⁹⁰ The authors presented the application of perovskite metasurfaces on optical encryption.⁹⁰ The perovskite metasurface also can be used in optical phase control, which was confirmed by Zhang et al. in 2019. They also used the EBL flowed ICP technique to prepare MAPbX3 cut-wire metasurfaces on metal substrates ^{[Figs. 6(i) and 6(j)]}.⁹⁶ They found that these MAPbX3 metasurfaces can generate a full phase control from 0 to 2π and high-efficiency and broadband polarization. Finally, they proved the potential application in holographic images based on the unique property of perovskite metasurfaces.⁹⁶

Besides 1D and 2D structured-perovskite, perovskite-based 3D structures have also been investigated. In 2017, a two-step method for the fabrication of CsPbX3 microcubes with subwavelength size was developed by Hu et al.⁹⁹ These CsPbX3 microcubes had a regular cube shape and smooth end faces, displaying tunable emission and excellent structure stability for several months under ambient conditions. In the same year, Zhang and coworkers used the CVD method on the prepared CsPbX3 MSs with controlled diameter of ∼1μm and tunable PL ranging from 425 to 715 nm ^{[Figs. 7(a) and 7(b)]}.¹⁰⁰ Wei et al.¹⁰¹ developed an automated microreactor system to fabricate an inorganic perovskite NCs sphere by UV photoinitiated polymerization in flow-focusing microfluidics ^{[Figs. 7(c) and 7(d)]}. These obtained CsPbBr3 spheres had a large diameter around 100μm, and the diameter could be influenced by flow rates.¹⁰¹ Mi et al.¹⁰² used the CVD method to fabricate high-quality single MAPbBr3 crystals with a cube-corner pyramids shape and lateral dimension in the range of 2 to 10μm on mica substrates ^[Fig. 7(e)]. Then, Yang et al.¹⁰³ also used the CVD method to fabricate CsPbI3 triangular pyramids with a spontaneous emission of ∼719nm at room temperature on a Si/SiO2 substrate ^[Fig. 7(f)].

Except for the regular morphologies, complex perovskite structures have also been investigated. Chen et al.¹⁰⁴ used a seed-mediated solvothermal method to fabricate monodisperse CsPbX3 NCs with nanoflower morphology [in Figs. 7(g)–7(i)]. Figure 7(h) shows the growth process of CsPbBr3 nanoflowers, which is formed by the structure transformation from Cs4PbBr6 to CsPbBr3. It is obtained that CsPbX3 dodecapods contained 12 well-defined branches, with a PLQY of about ∼50%. Moreover, the PL emission could be tuned from 415 to 685 nm. They prepared a white LED device based on using CsPbBr3 nanoflowers, exhibiting the 135% National Television System Committee (NTSC) standard.¹⁰⁴ In 2019, Li et al. fabricated single crystal microcuboid-MAPbBr3 and multistep-MAPbBr3 NCs via the solvothermal method at 120°C. In this process, microcuboid-MAPbBr3 was formed initially, and then the center of the surface was etched after long-time reaction, inducing the formation of multisteps. By adjusting the reaction temperature and time, the morphology and size of microcuboid-MAPbBr3 ^[Fig. 7(j)] could be adjustable, with performing potential in perovskite nanolaser and other optoelectronic devices.¹⁰⁵

Nonlinear optics describes the nonlinear state of the interaction between light and matter.^106–111 The researches on optical nonlinear materials are fundamental to nonlinear optics devices such as optical storage, optical switches, optical amplifiers, and lasers.^112–114 Due to the multiformity of physical and chemical properties, halide perovskites have been demonstrated as promising materials as nonlinear optics materials, which are related to the component and crystal structure of perovskite NCs.¹⁰⁶

In 2015, Sargent and coworkers investigated two-photon absorption in MAPbBr3 single crystals, under ultrashort pulses 800 nm excitation ^{[Figs. 8(a)–8(d)]}. They observed two-photon PL around ∼572nm with an absorption coefficient of 8.6±0.5cmGW−1 at 800 nm.¹¹⁷ Later, Heiko et al. performed temperature-dependent PL measurements on MAPbBr3 single crystals under 810 nm excitation. They observed obvious wavelength shifts of PLs with variable temperatures, which was attributed to discrete transitions between several stable crystalline phases of MAPbBr3 single crystals.¹¹⁵ In 2016, Kalanoor et al. studied the nonlinear optical responses of MAPbI3 films by the Z-scan technique, under nanosecond and femtosecond pulsed lasers. The nonlinear refractive index under femtosecond excitation was ∼69×10−12 and ∼34.4×10−9cm2/W for resonant nanosecond excitation, which was equivalent to conventional semiconductors.¹¹⁸ The Z-scan study of MAPbX3 (X = Cl, Br, I) perovskite film under the 800 nm, 40 fs pulse indicated that MAPbI3 films have a relatively large nonlinear optical coefficient compared with the MAPbCl3 and MAPbBr3 films.¹¹⁹ In the case of inorganic perovskites, Sun and coworkers discovered nonlinear optical properties of CsPbX3 NCs for the first time ^{[Figs. 8(e)–8(g)]}. They observed strong two-photon absorption from 9-nm-sized CsPbBr3 NCs, with a large absorption cross-section of ∼1.2×105GM.¹¹⁶ The nonlinear optical properties of CsPbX3 perovskite are highly correlated with their morphology. Jiang and coworkers¹²⁰ investigated nonlinear optical properties of CsPbBr3 NSs with a dependence on their thickness. When the thickness of CsPbBr3 NS was adjusted from ∼104.6 to ∼195.4nm, PL intensity increased nearly three times. They demonstrated that the two-photon absorption coefficient is inversely proportional to the thickness of CsPbBr3 NSs.¹²⁰ Krishnakanth et al. investigated nonlinear optical properties from nanocubes and NRs by Z-scan technology, under femtosecond 600, 700, and 800 nm lasers. They obtained large two-photon absorption cross sections of ∼105 GM and strong nonlinear optical susceptibility of ∼10−10esu in these films.¹²¹

The laser is a process of amplifying optical signals and generating high-intensity coherent light through stimulated radiation and is usually composed of three parts: energy pumping source, gain medium, and optical resonator. The amplification of the laser can be quantified as the resonance ability of gain media.^122,123 For gain media, the optical gain of the semiconductor is similar to the optical absorption, which is suitable for perovskite.^124,125 At the same time, optical losses are generated in optical cavities, which mainly come from nonradiative recombination, phonon scattering, edge scattering, and field leakage in the interface of cavities.¹²⁶ Perovskite materials have ultralow density, inducing high optical gain and low optical losses for resonance in perovskite, enabling promising potential in perovskite lasers with low threshold. The optical gain of semiconductors can be calculated by the variable-stripe-length measurement, which is related to the dependence of the amplified luminescence on the length of the slit width of the excitation. Xing et al.^23,124 performed variable stripe length measurements on MAPbI3 with a gain coefficient of ∼250cm−1, which was close to that of conventional semiconductor materials. The obtained optical gain coefficients of MAPbBr3 and MAPbCl3 were ∼300 and ∼110cm−1, respectively.^127,128 Liu et al.¹²⁹ demonstrated that the optical gain coefficient of CsPbBr3 nanocuboids can be calculated to be ∼502cm−1 under the 800 nm laser. Then, Zhao et al. reported efficient two-photon ASE from CsPbBr3 single crystals with a millimeter size and an optical gain of 38cm−1.¹³⁰

In the case of perovskite QDs without an external cavity, the amplification was generated from multiple scattering between QDs, enabling random fluctuations of lasing modes.¹²⁷ In 2015, Kovalenko and coworkers reported low-threshold ASE from colloidal CsPbX3 NCs with an optical gain coefficient of ∼450cm−1 and threshold of ∼5μJ/cm2.¹²⁷ In Figs. 9(a)–9(c), the ASE from CsPbX3 NCs could be tuned from 440 to 700 nm. Finally, they obtained random lasing from CsPbX3 films without the resonant cavity and whispering gallery mode (WGM) lasing using a silica sphere as the resonant cavity ^[Fig. 9(c)].¹²⁷ Besides, coating perovskite QDs onto an external cavity, Zeng and coworkers developed another method to form resonant cavities for perovskite QDs. They obtained enhanced random lasing from strong scattering in the perovskite/SiO2 composite with low threshold of ∼40μJ/cm2 ^{[Figs. 9(d)–9(f)]}.¹³¹ Similarly, Yang et al.¹³⁶ realized upconversion random lasing from FAPbBr3/A-SiO2 composites with a threshold of ∼413.7μJ/cm2. Liu et al.¹³⁷ obtained WGM and random lasing with a threshold of ∼430μJ/cm2 under 800 nm excitation by embedding CsPbBr3 QDs into a single silica sphere. In addition, microcapillary tubes can also be used to build WGM cavities for perovskite QDs. In 2015, Zeng and coworkers observed lasing emission from CsPbBr3 QDs by filling the CsPbBr3 QDs into a capillary tube, which acted as a WGM cavity for perovskite QDs film around the inner wall.¹³⁸ Later, stable two-photon pumped WGM lasing was realized by coupling CsPbBr3 and FAPbBr3 perovskite QDs into microtubules with thresholds of ∼0.8 and ∼0.31mJ/cm2 ^{[Figs. 9(g)–9(j)]}, respectively.^132,139

Besides the realization of perovskite QDs lasing-based silica sphere and microcapillary tube, the well-designed distributed Bragg reflector (DBR) can also be used to achieve a vertical-cavity surface-emitting laser (VCSEL).^133–135 In 2017, Zeng and coworkers first fabricated VCSELs with a sandwiched structure of DBR/CsPbBr3 QDs/DBR, which exhibited a low threshold ∼9μJ/cm2 directional output and favorable stability ^[Fig. 9(k)].¹³³ The lasing emission of CsPbX3-based VCSELs can be tuned in the visible light range.¹³³ In the same year, Huang et al.¹³⁴ fabricated CsPbBr3 QDs VECSLs with ultralow threshold of ∼0.39μJ/cm2 ^[Fig. 9(l)]. Organic hybrid perovskites-based VCSELs have also been performed. Chen and Nurmikko¹³⁵ developed FAPbBr3-based VCSELs by embedding FAPbBr3 solid thin films in two DBRs ^[Fig. 9(m)] with a threshold of ∼18.3μJ/cm2 under subnanosecond pulse excitations. They also demonstrated that the VCSEL device fabrication process can be applicable to flexile substrates, as shown in Fig. 9(m), which extended further practical applications for perovskite-based laser devices.¹³⁵ Most recently, Li et al.¹⁴⁰ fabricated a two-photon-pumped MAPbBr3 VCSEL by intergrading MAPbBr3 with DBR and Ag mirrors with a threshold of ∼421μJ/cm2, a Q factor of ∼1286, and a small divergence of ∼0.5deg.

Owing to the difference between the refractive index of perovskite material and air, the reflection can occur at the output interface easily, acting as optical reflector.^141,142 Hence, different from QDs, single perovskite crystals structures such as rods, wires, plates, cubes, and spheres can act as Fabry–Pérot (F-P) or WGM cavities by themselves, since the light can be confined in the resonant cavity with regular morphology and smooth end faces.¹²⁵ For the 1D NWs structure, light will propagate along 1D and form resonance between two end-facets.¹⁴³ Hence, perovskite NWs and NRs have been confirmed as potential structures in optoelectronic devices and nanoscale-integrated photonics due to their unique optical properties, such as highly coherent output and efficient waveguide effect.¹⁴³

Zhu et al.¹⁴⁴ demonstrated perovskite NW lasers using high-quality MAPbX3 NWs, which had a regular shape with rectangular cross section ^{[Fig. 10(a)]}. Tunable F-P lasing could be observed from single MAPbX3 NWs with low threshold of ∼0.22μJ/cm2 and Q factor of ∼3600 at room temperature ^{[Figs. 10(b) and 10(c)]}. In the same year, Xing et al.⁴⁹ realized F-P lasing from MAPbI3 NWs with rectangular morphology and length of ∼20μm. The obtained NW laser exhibited low threshold of ∼11μJ/cm2 and Q factor of ∼405, and the lasing wavelength could be tuned in the range of 551 to 777 nm.⁴⁹ In case of lasing from all-inorganic perovskite NWs, Yang and coworkers realized F-P lasing from CsPbBr3 NWs with a threshold of ∼5μJ/cm2 and a Q factor of ∼1009.⁵⁵ Fu et al. realized wavelength widely tunable F-P lasing from CsPbX3 NWs. The lasing wavelength could be tuned in the visible spectral region from 420 to 710 nm ^{[Figs. 10(e)–10(g)]}.¹⁴⁵ In 2017, lasing emission from triangular CsPbX3 micro/NRs with an ultrasmooth surface by the vapor-phase approach was reported. The obtained lasing could be tuned in the range from 428 to 628 nm, with low threshold of ∼14.1μJ/cm2 and high Q factor of ∼3500.⁷⁵ Efficient multiphoton pumped lasing in a wide excitation wavelength range (700 to 1400 nm) was realized.¹⁴⁷ Most of the single perovskite NWs mainly exhibited single-band lasing emission. In 2020, Tang et al. fabricated a single CsPbCl3−3xBr3x alloy NW via a solid–solid anion-diffusion process. They realized continuous F-P lasing in single as-prepared NWs, which could be tuned from 480 to 525 nm ^{[Figs. 10(i) and 10(j)]}.¹⁴⁶

Different from the F-P cavity formed by NWs/NRs, the perovskite 2D structure such as NPs will result in the WGM optical resonant cavity, which has a higher Q factor than the F-P cavity. In 2014, Zhang et al. first realized WGM lasing from MAPbI3 NPs with well-defined hexagonal and triangular shapes under femtosecond-pulsed laser excitation. The lasing wavelength was located at ∼780nm with a threshold of ∼37μJ/cm2 ^{[Figs. 11(a)–11(d)]}.⁸⁰ Liao et al.¹⁵⁰ obtained single-mode WGM lasing from single MAPbBr3 MDs peaked at ∼557.5nm with a threshold of ∼3.6μJ/cm2 and Q factor of ∼430. Liu et al. realized WGM lasing from MAPbI3 MP arrays with low threshold of ∼11μJ/cm2 and Q factor of ∼1210.¹⁵¹ Moreover, they observed single mode lasing by shortening the size of MPs.¹⁵¹ Qi et al.¹⁵² demonstrated that the threshold of the MPs laser decreases linearly depending on the later size, and the cavity mode density increases with the size. In 2019, WGM lasing from a triangular MAPbI3 perovskite NP with a lateral length of 27μm and thickness of 80 nm was realized at room temperature. The threshold of the WGM laser was ∼18.7μJ/cm2 and Q factor was ∼2600 ^{[Figs. 11(e)–11(i)]}.¹⁴⁸

As for the 2D all-inorganic perovskite-based laser, Zhang et al.¹⁴⁹ obtained WGM excitonic lasing from single-crystalline CsPbX3 NPs with micron-size length and subwavelength thickness ^{[Figs. 11(j)–11(m)]}. Multicolor lasing from 410 to 700 nm was realized in these NPs at room temperature ^{[Fig. 11(l)]}. The lasing threshold of the CsPbX3 NP was as low as ∼2.0μJ/cm2, and the linewidth of the WGM modes was ∼0.14 to 0.15 nm ^{[Fig. 11(m)]}.¹⁴⁹ Zheng et al.⁸⁶ demonstrated that CsPbI3 perovskite NSs possess WGM lasing under both one- and two-photon pumps with low-threshold-pumped excitation ^{[Figs. 11(n)–11(q)]}. The thresholds of lasing were ∼0.30 and ∼2.6mJ/cm2 under one- (470 nm) and two-photon (1200 nm) excitation, and the Q factors were ∼1489 and ∼1179, respectively, which is three times higher than the reported values of organic–inorganic lead halide perovskite NS. Most recently, Liu et al.¹⁵³ realized two lasing modes (F-P and WGM) in the all-inorganic perovskite CsPb2Br5 MPs with subwavelength thickness and uniform square shape under two-photon pump. Remarkably, low-threshold F-P multimode lasing with Q factor of ∼3551 and single-mode WGM lasing with Q factor of ∼3374 from the same MP at room temperature have been achieved successfully.

A single perovskite spherical 3D structure has also usually been demonstrated as a WGM cavity. In comparison with other nano/microstructure resonant cavities, the coupling between the sphere cavity and substrate was relatively weak, which resulted in less optical losses. Zhang and coworkers realized single-mode lasing in CsPbX3 MSs with regular sphere shape and submicron size at room temperature ^{[Figs. 12(a)–12(d)]}.¹⁰⁰ The line width of WGM lasing was ∼0.09nm, the threshold was ∼0.42μJ/cm2, and Q factor was ∼6100 ^{[Fig. 12(c)]}. In addition, the single-mode lasing can be tuned in the whole visible region through element modulation and size control of perovskite MSs ^{[Fig. 12(d)]}.¹⁰⁰ Furthermore, they achieved two-photon single-mode lasing with linewidth of ∼0.037nm and Q factor of ∼1.5×104 from a single CsPbBr3 MS at room temperature, which are the best values obtained in perovskite-based micro/nanocavities until now.¹⁵⁴ Moreover, these perovskite MS lasers showed uniform lasing emission, which could be observed in the range from −30deg to 30 deg.^154,155

Another 3D structure generally used for perovskite lasing is the nano/microcube. Liu et al.¹²⁹ obtained F-P lasers from an individual CsPbBr3 nanocuboid with subwavelength scale for the first time ^{[Figs. 12(e)–12(h)]}. They realized single-mode F-P lasing from a CsPbBr3 nanocuboid with low thresholds of ∼40.2 and ∼374μJ/cm2 and Q factors of ∼2075 and ∼1859 under one- and two-photon pumps, respectively.¹²⁹ The physical volume of the obtained laser is ∼0.49λ3. Moreover, the pulse duration is only ∼22ps, which is consistent with the resulting fast decay of SE observed by fs transient absorption spectroscopy ^{[Fig. 12(h)]}.¹²⁹ Cube-corner pyramid cavities could also act as microretroreflectors. In 2018, Mi et al. realized F-P lasing in cube-corner MAPbBr3 pyramids at room temperature ^{[Figs. 12(i)–12(m)]}.¹⁰² Furthermore, the threshold of lasing could be reduced from ∼92 to 26μJ/cm2 by coating a thin layer of Ag film on a mica substrate ^{[Figs. 12(l) and 12(m)]}.¹⁰² Most recently, Yang et al.¹⁰³ also realized F-P lasing from a single CsPbI3 triangular pyramid with a microsize at low temperature. They demonstrated that the temperature-dependent lasing threshold can be reduced from ∼53.15 to 21.56μJ/cm2 with corresponding temperature from 223 to 148 K.¹⁰³

In comparison with single perovskite lasers, laser arrays with high-density patterns and high-precision arrangements are more necessary for mass-produced, compact on-chip optoelectronic circuit integration. In 2016, Wang et al. fabricated MAPbBr3 microwire arrays and realized high density perovskite lasers from these microwire arrays ^{[Figs. 13(a) and 13(b)]}, in which all of the subunits generated the same lasing emission.¹⁵⁶ The minimum unit period was 800 nm, presenting an integration density of nanolasers as high as 1250mm−1. In 2017, Fu and coworkers prepared MAPbBr3 NW arrays with the width from 460 to 2500 nm, height from 80 to 1000 nm, and length from 10 to 50μm. These perovskite NW arrays were demonstrated as almost identical optical resonance cavities with a low threshold of ∼10.2μJ/cm2 ^{[Figs. 13(c) and 13(d)]}.¹⁵⁷ In 2016, Liu et al. realized WGM lasing from patterned MAPBI3 microplatelets arrays with a threshold of ∼11μJ/cm2 and Q factor up to ∼1210. The wavelength tunability and single mode lasing could be selected by changing platelet sizes or breaking the symmetry of the designed laser pattern.¹⁵¹ In the same year, Feng et al.¹⁵⁸ demonstrated that the MAPbBr3 square MP array shows high-performance WGM lasing with tunable mode and low lasing threshold ^{[Figs. 13(e) and 13(f)]}. Single mode lasing was obtained from a 2.1-μmMAPbBr3 square. Lin et al.¹⁵⁹ fabricated a large-area CsPbX3 QDs array by a photolithographical approach, which could be used as efficient lasing structures and emitting pixel arrays ^{[Figs. 13(g)–13(i)]}. They realized WGM lasing from the QD arrays with high Q factor and demonstrated that this patterning technique can be used in large-area perovskite laser arrays with multicolor pixels ^{[Fig. 13(g)]}.¹⁵⁹ Most recently, Wang et al.¹⁶⁰ fabricated large-area MAPbX3 MD arrays via a screen-printing technique ^{[Figs. 13(j)–13(m)]}. They obtained tunable WGM lasing from these MAPbX3 MD arrays with a threshold of ∼21.3μJ/cm2 and a Q factor of ∼1570 successfully. Multicolor WGM lasing emission could be tuned from ∼510 to 650 nm ^{[Fig. 13(l)]}.¹⁶⁰ In 2020, Song and coworkers employed the topologically protected optical bounded states in the continuum (BICs) and demonstrated the ultrafast control of perovskite-based vortex microlasers at room temperature. They proved that vortex beam lasing based on perovskite metasurfaces could be switched to linearly polarized beam lasing with switching time of 1 to 1.5 ps. The energy consumption was several orders of magnitude lower than that of previously reported all-optical switching.¹⁶¹

Surface-plasmon (SP) is an excited state with large enhancement of the electromagnetic field localized at the metal–dielectric interface, which provides confinement on the subwavelength scale, overcoming the diffraction limit of light.¹⁶² In perovskite micro/nanolasers, SPs have been demonstrated as an effective method to tailor the properties of lasers. In general, SPs could be generated by the metal layer, such as Au or Ag, and transfer along the semiconductor–metal interface. Kao et al.¹⁶³ reduced the lasing threshold of perovskite by strong exciton–plasmon coupling between the Ag and perovskite films ^{[Fig. 14(a)]},¹⁶³ in which the confined optical fields between Ag and perovskite films could be enhanced about 19.3 and 7.7 times in comparison with bare perovskites and perovskites coated by Ag thin film, respectively. In 2017, Wang et al. deposited Al nanoparticles onto the surface of CsPbBr3 perovskites. The lasing thresholds of CsPbBr3 perovskite microrods were significantly reduced by >20%, and the output intensities were significantly enhanced via the plasmonic resonances.¹⁷² In 2019, Wu et al. reported a method to enhance ASE performance of MAPbI3 films by adding Au NRs-doped PMMA on MA3PbI3 perovskite films. The ASE threshold was significantly reduced by ∼36%, and the output intensity increased by 13.9-fold with the plasmon resonance enhancement of Au NRs.¹⁷³ Yang et al.¹⁷⁴ also reduced the lasing threshold of CsPbBr3 perovskite nanocubes significantly by ∼33% via the surface plasmonic effect of Au nanoparticles. In 2021, single-mode upconversion plasmonic lasing from MAPbBr3 perovskite NCs was realized by Lu et al.,¹⁶⁴ exhibiting low threshold ∼10μJ/cm2 and small mode volume ∼0.06λ3 at 6 K, where TiN was used as a promising resonance adjustable plasmonic platform ^{[Fig. 14(b)]}. Hsieh et al.¹⁶⁵ realized continuous-wave (CW) lasing from a single CsPbBr3 QD in a plasmonic gap-mode nanocavity with low threshold of ∼1.9W/cm2 and small mode volume of ∼0.002λ3 ^{[Fig. 14(c)]}. Most recently, Li et al.¹⁶⁶ proposed a hybrid nanocavity composed of CsPbBr3 nanoparticles and a thin Au film, which could realize optically controlled quantum size effect by the reversible phase transition from polycrystalline to monocrystalline ^{[Fig. 14(d)]}. These results demonstrated that SPs could not only modulate the performance of perovskite lasers but also can realize deep subdiffraction plasmonic lasers.

Perovskite is also an ideal candidate to realize a room temperature exciton polariton laser, which mainly results from the strong exciton–photon coupling between the gain media and nanocavity. In perovskite laser researches, room temperature exciton polaritons have been realized with various nanostructures.^{167,168,175,176} Perovskite NCs with self-assembled morphology can provide optical resonators due to the confinement of exciton–photon coupling. On the other hand, a planar optical cavity composed of two mirrors can be used as an F-P cavity conventionally. In 2018, Liu et al.¹⁶⁷ observed strong exciton–photon coupling in single CsPbBr3 micro/NWs and MAPbBr3 micro/NWs, respectively ^{[Fig. 14(e)]}. Moreover, polariton lasing was realized at room temperature with exceptionally large vacuum Rabi splitting of ∼656 and 390 meV.^167,175 Shang et al.¹⁷⁶ proved light could propagate as an exciton–photon in CsPbBr3 NWs at room temperature, increasing optical absorption and emission in comparison with bulk crystals. They demonstrated that the decrease of CsPbBr3 dimensions could enhance the exciton–photon coupling strength, which increased the exciton fraction. Furthermore, they found that the increase of temperature could significantly decrease the exciton fraction of exciton–photons, causing high thresholds and restraining CW lasing above 100 K. They successfully realized CW-pumped lasing from CsPbBr3 nanoribbons by reducing the height to ∼120nm on sapphires with low threshold of ∼0.13kW/cm2 ^{[Fig. 14(f)]}.¹⁶⁸ Then, they coupled MAPbBr3 NWs with a hybrid plasmonic microcavity to enhance exciton–photon interaction.¹⁷⁷ They observed a Rabi-splitting up to ∼564meV in a hybrid perovskite/SiO2/Ag waveguide microcavity at room temperature. In 2017, Su et al.¹⁷⁸ reported room-temperature polariton lasing based on an epitaxy-free all-inorganic CsPbCl3 nanoplatelet embedded in DBRs, supporting F-P oscillations. The polariton lasing exhibited a threshold of ∼12μJ/cm2. Zhang et al.¹⁶⁹ investigated the trapping of polaritons in micron-sized CsPbBr3 flakes embedded in DBRs as a microcavity ^{[Fig. 14(g)]}. They demonstrated quantized polariton states arising from the optical confinement of flakes.

In comparison with perovskite NCs with 3D structure, quasi-2D perovskites have a quantum well (QW) structure with the advantages of large exciton binding energy and low nonradiation loss and are more easily coupled with a resonant cavity. In 2018, Fieramosca et al. observed strong exciton–photon coupling from hybrid 2D perovskite flakes. The organic ligands efficiently affected the out-of-plane exciton–photon coupling, suggesting that the organic interlayer plays a significant role in the anisotropy of the exciton and exciton polariton.¹⁷⁹ Then, they observed highly interacting polaritons in (PEA)2PbI4 with an excitonic interaction constant as ∼3μeVμm2, which was two orders higher than that of organic excitons.¹⁸⁰ Zhang et al.¹⁸¹ investigated cavity polariton modes in 2D perovskite (PEA)2PbBr4 sheets. The perovskite layer naturally could act as an F-P cavity and exhibited evident cavity polariton modes with Rabi splitting energy of ∼259meV. Li et al.¹⁷⁰ first reported room temperature optical gain from 2D perovskite (NMA)2FAn−1PbnX3n+1 (NMA=C10H7CH2NH3+). In these layered perovskite nanostructures, multiple QW phases naturally form an energy cascade, enabling an ultrafast energy transfer process from higher energy bandgap QWs (n<5) to lower energy bandgap QWs (n>5). They obtained tunable ASE ranging from 530 to 810 nm with low ASE threshold (<20.0μJ/cm2) ^{[Fig. 14(h)]}. Later, lasing based on these quasi-2D perovskite nanostructure has also been realized by researchers, e.g., Liang et al. investigated multicolor lasing from (BA)2(MA)n−1PbnI3n+1 (BA=C4H9NH3+) in 2019.¹⁸² Most recently, Liu et al. shrank the quasi-2D perovskites laser to the deep-subwavelength scale with 50 nm, which was the smallest room temperature all-dielectric laser.¹⁸³ They revealed the contribution from excitons and polarons to the high optical gain, which provided an insight into the design of next-generation integrated laser sources. Qin et al.^171,184 found that the triplet excitons in hybrid quasi-2D perovskite have a lifetime up to 1μs, which might cause the disappearance of the laser. Then, using a distributed-feedback (DFB) cavity with a high Q and triplet management strategies, they realized stable room-temperature CW lasing in quasi-2D perovskite films ^{[Fig. 14(i)]}. The representative works about perovskite lasers in recent years are summarized in Table 1. All of these progresses prove the potential of perovskite materials in micro/nanolasers.

Over the last few years, tremendous investigations have been carried out on metal halide perovskite materials, especially studies of the corresponding physicochemical properties and exploration of relevant applications in optoelectronic devices. In this review, we summarized the recent developments of the synthesis strategies, the morphological control, and lasing application of metal halide perovskite materials. The various synthetic methods for the fabrication of perovskite NCs have been investigated in previous researches, including the solution method and chemical deposition method. Moreover, the morphology of perovskite NCs can be controlled with different dimensions via adjusting the reaction conditions. Their structure-related optical properties were investigated on the single-particle with various structures as 0D, 1D, 2D, and 3D, enabling their potential in LEDs, solar cells, photodetectors, and lasers.

In spite of the tremendous advances in perovskite materials and perovskite-based lasers so far, there are still many issues to be further solved. The central issue of perovskite materials is their instability, which is the biggest obstacle for their industrialization. Although enormous work has been performed to enhance the stability of perovskites, such as the surface ligand modification or encapsulation method, the instability characteristic of the perovskites still limits their commercial applications. So far, the mechanisms of their decomposition are not yet clearly understood, hindering their further performance improvements. Another important issue related to lead halide perovskite materials is the urgent trend of reducing or removing the lead element due to its toxicity. For this purpose, some strategies have been proposed to constitute lead-free perovskites by possible substitutes using either homovalent elements such as Sn and Ge or heterovalent elements such as Bi and Sb.^185,186 Unfortunately, the optoelectronic properties of lead-free perovskites have not been effectively improved. Furthermore, the nucleation and growth mechanisms of perovskite NCs are yet to be revealed clearly, which is helpful to accurately control the morphology of the perovskite NCs for better understanding structure–property relationships. Last, but not least, theoretical explanation about the photophysics of perovskite NCs is necessary to better explain the quantum size effects of perovskite crystals, which could guide the research directions to regulate and control their electronic, optical, and defect properties.

The potential of perovskite materials in laser applications has been abundantly demonstrated. We reviewed a variety of laser cavities and summarized the dependence between the resonant cavity and the structure of perovskite NCs. Various linear and nonlinear perovskite lasers with an ultralow threshold have been realized in single perovskite NCs with different dimensions. Owing to the large gain coefficient and long-distance ambipolar carrier-transport, perovskites have great potential in electrically driven lasers, which have huge application value in integrated optoelectronic devices. But until now, all of the obtained perovskite lasers are pumped by laser excitation. The research about electrically pumped perovskite lasers has not been realized. Further investigation of resonant cavities, together with further reduction of the lasing threshold under optical excitation via optimization of the material properties, will boost the realization for electrically driven lasers of perovskites. The future trend of the perovskite-based laser is to integrate with optoelectronic components for further waveguide and signal processing. More importantly, the resonance and gain of perovskite materials and perovskite-based lasers need photophysical theory, which will inspire exploring the carrier relaxation and charge transfer processes of high-performance devices.