The laser technique has been successfully used for producing high-power, ultra-fast, and small coherent light sources. The term “laser” stems from an acronym for “light amplification by stimulated emission of radiation”, which can emit coherent light with a strong intensity and superior directionality through a process of optical amplification based on the stimulated emission of electromagnetic radiation.^[1–4] Due to the rapid development of nanoscience and nanotechnology, small solid-state lasers, with physical size comparable to the optical wavelength, have been extensively utilized in a variety of fields, such as high-density data storage, optical integration, and high-resolution bio-imaging. ^[5,6] Inorganic and organic semiconductor nanostructures in the form of nanowires (NWs) and nanoplatelets (NPs) have been considered as active optical microcavities for achieving lasing over the past three decades.^[7] Along with technological innovation, combining gain materials with optical cavities (such as photonic crystals or metal-clad cavities) has been deemed a vital method to realize photonic lasing. For instance, Chen et al. successfully realized the single mode, low threshold, and spatially coherent laser operation in the near-infrared region by coupling a semiconductor thin film with photonic crystals.^[8] Nevertheless, due to the diffraction limit, the mode and physical dimension of the photonic laser are larger than half of the wavelength of the optical field, which triggers a crucial fundamental challenge to realizing ultra-compact lasers that can directly generate coherent optical fields on a nanometer scale.^[9,10] The plasmonic laser, a new class of optical amplification and laser device, has been demonstrated to store electromagnetic field energy in charge density waves.^[11,12] A hybrid semiconductor–insulator–metal construction is widely adopted to overcompensate for the losses to realize plasmonic laser, which will be discussed in detail in Section 4.^[13] Furthermore, polariton lasers have attracted great attention and play an irreplaceable role in the field of lasing. Unlike photonic and plasmonic lasers, polariton lasers have no threshold condition linked to the inversion of the population and also open up a host of technological possibilities in the applications of slow light, nonlinear optics, and low-threshold coherent emitters.^[14,15]

Metal halide perovskites with the general formula ABX₃ or A₂BX₄ are considered as the most promising materials for future optoelectronics. In 2009, Miyasaka et al. first used the CH₃NH₃PbI₃ (MAPbI₃) as a light absorber in solar cells with an initial efficiency of only 3.8%, Then, due to the improvement of materials and interface engineering, the efficiency was enhanced to over 22%.^[16–23] The high absorption coefficient, low nonradiative recombination rate, long diffusion length, low density of defect states, and bipolar carrier-transport property make the perovskites family uniquely suitable for photovoltaic applications.^[18,24–26] Besides, owing to the outstanding emission properties such as high photoluminescence (PL) quantum yield (QY) and widely-tuned band gap, perovskites have been extensively utilized in the field of optoelectronic devices, such as in light emitting diodes and photo-detectors.^[27–29] Moreover, by virtue of high optical gain, perovskites are regarded as an efficient optical gain medium for lasing applications, including photonic lasing, plasmonic lasing, and polariton lasing. For example, Xing et al. first demonstrated the amplification of spontaneous emission (ASE) from a low temperature solution-processed MAPbX₃ perovskite thin film in 2014.^[30] In addition, through laying MAPbI₃ NWs on Ag film with a MgF₂ spacer layer, Yu et al. demonstrated Fabry–Pérot (F–P) plasmonic lasers in perovskites.^[31] Furthermore, Su et al. realized polariton lasing based on all-inorganic cesium lead chloride (CsPbCl₃) perovskite NPs at room temperature.^[32] In a nutshell, on the basis of the advantages of low-cost fabrication, perovskite microlasers and nanolasers have crucial potential applications in integrated optics. Hence, we will give a general overview about the rapid process of perovskite-based small laser sources. This paper is mainly divided into six parts to review the current research status of perovskite-based small lasers. In Section 2, we will describe the structure and optical property of perovskites. In Section 3, we will review the photonic lasers based on diverged reduced-dimensional perovskites, including NPs, NWs, quantum dots (QDs), and thin films. In Sections 4 and 5, we will briefly discuss the plasmonic laser and polariton laser, respectively. Finally, we will provide a conclusion and give perspectives on future research areas in perovskite small lasers.

Hindered by the optical diffraction limit, the physical size of a photonic laser is constrained above half of the light wavelength. Nanolaser devices based on the amplification of surface plasmons by stimulated emission of radiation are utilized to push the dimensions of laser devices into the nanometer region beyond the diffraction limit.^{[11,13,81–86]} Oulton et al. proposed a hybrid plasmonic waveguide for subwavelength confinement and long-range propagation, which consists of a semiconductor NW separated from a metal surface by a nanoscale dielectric gap. Owing to the continuity of the electric displacement near the dielectric spacer layer, the mode of perovskite NW is confined to an ultra-small mode volume in the dielectric gap region. The coupling between the plasmonic and waveguide modes across the gap makes it possible to store optical energy like a capacitor that allows effective subwavelength transmission in a nonmetallic region with low loss, as shown in Fig. 6(a).^[87] On this basis, a deep subwavelength mode confinement of λ²/400 to λ²/40 can be obtained, and more importantly, surface plasmon polaritons can propagate large distances of 40–150 μm. Based on the structures, plasmonic lasing in the nanometer scale has been realized separately in CdS, GaN, and ZnO NWs.^[13,88,89] Recently, Yu et al. demonstrated F–P plasmonic lasers based on perovskite NWs. Figure 6(b) shows a schematic of the plasmonic nanolaser: a single MAPbI₃ NW sitting on a Ag film with a 10 nm MgF₂ spacer layer. Figure 6(c) shows the calculated cross-sectional electric field distribution and polarization direction of the hybrid plasmonic mode at the emission wavelength of perovskite (790 nm). The optical feedback is provided by the two end facets of the NW, which can serve as mirrors to form an F–P cavity for forming the plasmon lasing.^[31] Compared with the photonic NW laser, where a NW is placed on quartz directly, the plasmonic laser can support optical modes of a free diffraction limit on a nanometer scale where photonic modes cannot even exist. As indicated by the vertical dotted line in Fig. 6(d), the calculated fundamental mode of photonic lasing goes through a cut off point at the diameter of about 146 nm. However, the fundamental mode of hybrid plasmon lasing does not experience any cut off point, which further demonstrates the superiority in this respect. Figure 6(e) shows the plasmonic lasing emission spectra varying with pump power. The Q-factor of the sample here is evaluated to be 151, and the Purcell factor is estimated at 8.3, showing good performance of the lasing characteristics.^[31] The research on perovskite plasmonic lasing is still quite limited, and the interaction between perovskite and plasmonic structures is rarely explored. Actually, the underlying photophysics in the new emergent perovskites is not well understood. The hybridization of perovskite and plasmonic structures, i.e. how to use plasmonic structure to tune the exciton properties and therefore reduce the optical losses, is an interesting topic, and more technologies can be generated for lasing applications accordingly.

Fig. 6.

Figure Option ViewDownloadNew Window

Fig. 6. (color online) F–P NW plasmonic lasers. (a) Hybrid plasmonic waveguide.[87] (b) Schematic of plasmonic nanolaser: a single CH3NH3PbI3 (MAPbI3) NW on Ag film with a MgF2 spacer layer.[31] (c) Calculated NW mode distribution for MAPbI3 NW on Ag film with a 10 nm MgF2 spacer layer at a wavelength of 790 nm.[31] (d) Calculated effective mode index against NW height for plasmonic and photonic lasing geometry.[31] (e) Emission spectra for different pump intensities, with the inset showing an optical image of the plasmonic lasing NW.[31]

Comparing with photonic and plasmonic lasing, the concept of polariton lasing was first proposed theoretically in 1996. Unlike photonic lasing accomplished by light amplification through the process of stimulated emission with population inversion, polariton laser based on the Bose–Einstein condensation (BEC) of exciton–polariton is a type of coherent emission without population inversion.^{[14,32,90–92]} In semiconductor microcavities, exciton–polaritons are formed when the interaction time between exciton and polaritons is shorter than their escaping time, which is the so-called strong interaction regime. As shown in Fig. 7(a), the most widely used microcavities for polariton lasing are similar to vertical cavity surface emitting lasing consisting of a planar F–P resonator embedded in two reflectors (usually DBR mirrors) and quantum wells or 2D NPs as the source of excitons. Excitons, as bosons, can couple with cavity photons to form new half-light, half-matter bosonic quasiparticles called polaritons whose effective mass is around 10⁴ times less than that of free electrons. According to theoretical analysis, the small effective mass is beneficial to the realization of BEC. Strongly coupled exciton–polaritons then generate the upper polariton branch (UPB) and the lower polariton branch (LPB) in the dispersion relation of energy. Considering that the lifetime of polaritons in the strong coupling regime is extremely short, the whole process cannot reach thermal equilibrium; however the relaxation time is also very short to ensure the final production of thermalized polariton gas.^[90]

Fig. 7.

Figure Option ViewDownloadNew Window

Fig. 7. (color online) Polariton lasing based on perovskites. (a) Microcavity consisting of a planar F–P resonator embedded in two reflectors. Excitons in an active layer and cavity photons can strongly couple with each other to generate new quasiparticles, exciton–polaritons. (b) Mechanism of ideal polariton lasing. Under high-energy excitation, polaritons are created and then leave the UPB for the LPB via phonon emission. Next, they leap down mainly through polariton–phonon scattering before reaching the bottleneck region and through polariton–polariton scattering after reaching the bottleneck to relax continuously to the final state k∥ = 0.[90] (c) Energy–wave vector (E–k) dispersion curves of CH3NH3PbBr3 (MAPbBr3) micro/nanowire with a width of 0.32 μm and length of 3.66 μm. The L–T splitting energy (ΔELT) is about 33 meV, which shows the strong coupling strength. The inset shows a normalized electric field distribution |E|2 at the cross-section of each corresponding wire.[15] (d) Lasing emission spectra (green line) and spatially resolved PL spectra (blue line) of a MAPbBr3 wire in the same condition. The red dots displayed in wave vector space with integer values of π/Lz are F–P peaks extracted from spatially resolved PL spectra, which can be fitted with the polariton dispersion curve (red line).[15] (e) Schematic diagram of the microcavity structure, where CsPbCl3 NPs are embedded in two DBR mirrors consisting of 7 and 13 HfO2/SiO2 pairs, respectively.[32] (f) Angle-resolved PL spectrum at high pump fluence of 1.3Pth. Almost all areas around the final state k∥ = 0 are occupied, indicating the realization of the whole polariton condensation.[32]

Figure 7(b) shows the mechanism of ideal polariton lasing without population inversion.^[90] There are two different methods to achieve polariton production: nonresonant excitation at k_∥ = 0, and resonant excitation at a larger angle. Under external high-energy excitation, i.e. from an optical pump, a mass of polaritons are created and later fall into the LPB from the UPB through phonon emission. After occupying the exciton-like reservoir in the LPB, polaritons continue to relax mainly by phonon emission until they reach the theoretical value at k_∥ = 0. However in this process, phonons with a bigger momentum sharply reduce the probability of emission, causing the relaxation to be not in one step and polaritons to have to “jump down” via losing tiny amounts of energy each time. Due to the fact that phonon–polariton scattering becomes difficult and the lifetime of polaritons decreases close to that of the photon-like final states, the bottleneck effect will occur, which has been observed in several experiments. For further condensation, polariton–polariton scattering is considered as the primary interaction to realize polariton population accumulation in the minimum energy state. Specifically, two polaritons stuck in the bottleneck can exchange energy with each other to make one leap into the area near k_∥ = 0, and the other polariton regains a higher momentum. Then, the latter repeats first-stage relaxation once more, eventually inducing polaritons to exclude the bottleneck region as a whole. In the ensuing discussion, we can see, via image analysis, two kinds of scatterings leading to relaxation successively. Finally, the condensate polaritons emit bright coherent light through their photonic component, similar to the scenario of photonic lasing, which can explain the origin of polariton lasing. It is worth mentioning that polariton emission usually shows a continuous blueshift due to the loss of strong coupling under higher pumping or the excitonic part of polaritons under lower pumping, which has been reported in different polariton structures.^[14]

In light of the above, strong exciton–photon coupling is the prerequisite for realizing polariton lasing. Over the past two decades, polariton lasing has been demonstrated in inorganic and organic semiconductors. In the earliest research, CdTe, GaAs, and InP have fulfilled potential in polariton condensation below 70 K due to small exciton binding energy. In a subsequent study, room temperature exciton–polariton lasing was achieved in a ZnO- and GaN-microcavity with much higher oscillator strengths and exciton binding energy.^[90,93–97] For organic semiconductors, generally speaking, Frankel excitons exhibit larger exciton binding energy but weaker Coulomb interaction, which is against polariton relaxation. In view of the facts presented above, the perovskite system can combine the advantages of both organic and inorganic materials, such as large exciton binding energy, easy fabrication, low trapping states density, and inter-system transition, to achieve high-performance polariton lasing devices.^[98–101]

The exciton–polariton has been observed in the system of perovskites, such as (C₆H₉C₂H₄NH₃)₂PbI₄, (C₆H₅C₂H₄NH₃)₂PbI₄, and CsPbX₃. Figure 7(c) shows the conventional energy–wave vector (E–k) dispersion of one MAPbBr₃ NW with width × length dimensions of 0.32 μm × 3.66 μm. Meanwhile, the obvious anticrossing feature indicates the occurrence of strong coupling, just like in previous research results. The longitudinal–transverse (L–T) splitting (ΔE_LT) is an index of coupling strength, and the L–T splitting energy here is 33 meV, which increases with the decrease in perovskite microcavity size.^[15] Another important parameter, the vacuum Rabi splitting energy in MAPbBr₃ is ∼ 390 meV, describing a large oscillator strength. Nevertheless, subsequent experiments prove that the lasing here should be photonic lasing rather than polariton lasing. In Fig. 7(d), the lasing peaks are not located in the final state above the threshold, and polariton condensation does not finish, probably due to the low efficiency of relaxation in the bottleneck region. Besides, the continuous blueshift at higher power is almost not visible, suggesting that it is not polariton lasing. Further research remains to be done, and they do offer the possibility of polariton lasing applications in perovskite systems.^[15]

Although room temperature, strong exciton–photon coupling in perovskite material has been observed over the past twenty years, the realization of polariton condensation is still a recent breakthrough. In 2017, Su et al. first reported polariton lasing in CsPbCl₃ NPs at room temperature.^[32] The all-inorganic lead chloride perovskites were grown by epitaxial-free CVD and then dry-transferred to a bottom DBR composed of 13 HfO₂/SiO₂ pairs. Through standard e-beam evaporation, seven other pairs were deposited continuously over NPs to form a complete DBR microcavity structure (Fig. 7(e)). The thicknesses of HfO₂ and SiO₂ were precisely designed to support F–P oscillations, and due to the high crystalline quality, the Q-factor of the optical microcavity was estimated to be 300. In order to verify the existence of polariton lasing in the system, the method of different off-resonant excitations was utilized to the probe power-dependent angle-resolved PL spectrum. As the pump fluence increases, the ground state emission intensity at k_∥ = 0 shows a transition from a linear region to nonlinear region at the threshold point (P_th = 12 μJ·cm⁻²), the FWHM narrows rapidly, and the intensity increases sharply by more than three orders of magnitude, just like the behaviors in traditional photonic lasing. Under the lower pumping condition, the LPB dispersion exhibits an average emission distribution at all angles, and the emission intensity at k_∥ = 0 becomes dramatically stronger than the others corresponding to polariton condensation upon approaching the threshold fluence. When the excitation power reaches 1.3P_th, almost only the ground state near the minimum of the LPB dispersion is massively occupied, whose energy is ∼ 120 meV lower than that of the minimum of the uncoupled cavity mode, indicating the difference between polariton lasing and photonic lasing shown in Fig. 7(f). In the process of polariton condensation, electron–hole pair relaxation is mainly connected with the polariton–reservoir (in fact the excitons in the reservoir) interaction and polariton–polariton interaction. They can lead to two distinct polariton emission energy blueshift trends, which are observed to be consistent with the theoretical results. Meanwhile, polariton condensation emerges clearly.

Finally, we summarize polariton lasing in perovskites. Strong exciton–photon coupling and polariton condensation are the latest concepts, and related studies are still underway. Light–matter interactions in perovskites have been proved constantly, however, in addition to large coupling strength and high polariton density, the realization of a complete relaxation process is also a key points. Devices based on DBR microcavities or NW F–P cavities are feasible, and how to improve the cavity quality will be the next problem for scientists to achieve novel polariton lasing. As a matter of fact, when under higher pumping, polariton lasing can transform into photonic lasing; hence, this also indicates a better future for polariton lasing.

In this review, we have taken a holistic view to introduce the current status and rapid process of micro-sized and nano-sized lasers based on metal halide perovskites. According to the quantum particles involved in the coherent emission process, lasers are divided into three categories, i.e. photonic lasers, plasmon lasers, and polariton lasers. Due to the excellent properties of high absorption and gain coefficient, low density of trapping states, and tunable emission, metal halide perovskites promise to be used for low-cost, multicolor lasing devices. This is a good opportunity for developing polariton lasing by using perovskite as the gain medium, which is attributed to their large exciton binding energy.

However, challenges still remain. The long-standing problem of poor structural stability due to ionic chemical bonds casts a shadow on practical applications. The 2D structure perovskites exhibit higher stability than their 3D counterparts. The 2D and 3D hybrid structure perovskites provide a solution to the stability issue. Electrical and CW optical pumping is another central challenge to making small perovskite lasers for industrial applications. Although the outstanding electric properties of perovskites have been proven widely in the solar cell research field, small, electrically driven perovskite lasers are rarely explored. The reason for this may lie in the instability of perovskite exposed to air, heat, and organic solvents during the fabrication of microdevices. The high lasing threshold is the other obstacle for electrically driven lasers. Although polariton lasing can reduce the threshold by two–three orders theoretically, the threshold of the as-demonstrated polariton lasers is still quite high. By using MAPbI₃, the first CW pumped lasing was developed due to local gain enhancement near the phase transition temperature. However, the underlying physics and universality of the method remain to be confirmed. Interdisciplinary studies on structural engineering, device fabrication, and fundamental photophysics will be helpful for the application of small perovskite lasers.