Transport Layer Engineering Toward Lower Threshold for Perovskite Lasers

Charge‐transport layers are essential for achieving electrically pumped perovskite lasers. However, their role in perovskite lasing is not fully understood. Here, the role of charge‐transport layers on the lasing actions of perovskite films is explored by investigating the amplified spontaneous emission (ASE) thresholds. A largely reduced ASE threshold and enhanced ASE intensity is demonstrated by introducing an additional hole transport layer poly(triaryl amine) (PTAA). It is shown that the key role of the PTAA layer is to accelerate the hot‐carrier cooling process by extracting holes in perovskites. With reduced hot holes, the Auger recombination loss is largely suppressed, resulting in decreased ASE threshold. This argument is further supported by the fact that the ASE threshold can be further reduced from 25.7 to 7.2 µJ cm−2 upon switching the pumping wavelength from 400 to 500 nm to directly avoid excess hot‐hole generation. This work exemplifies how to further reduce the ASE threshold with transport layer engineering through hot‐hole manipulation. This is critical to maintaining the excellent gain properties of perovskites when integrating them into electrical devices, paving the way for electrically pumped perovskite lasers.


Introduction
Electrically driven surface-emitting lasers, with an output of a coherent light source, are a fast-growing technology in various applications such as information transmission, medical treatment, and 3D sensing. [1][2][3] However, the current successful industrial fabrication of surface-emitting lasers is only based on a few inorganic materials, [4,5] heavily limited by the expensive and complex high vacuum manufacturing processes. Recently, metal halide DOI: 10.1002/adma.202300922 perovskites emerge as a promising candidate for future lasers owing to various advantages such as low-cost solution processability, bandgap-tunable luminescence with high color purity and high photoluminescence quantum yields (PLQY), outstanding optical gain coefficient, and excellent optoelectronic properties. [6,7] With intensive research interests, two significant breakthroughs have been made: the achievement of room temperature optically pumped continuous wave (CW) lasers, [8] and the realization of high injection current densities (>1 kA cm −2 ). [9] These two breakthroughs make halide perovskites very promising for electrically pumped lasers.
Toward demonstrating electrically pumped lasers, it is necessary to sandwich the perovskite thin film into transport layers for the integration of electrical devices. However, it remains unclear how transport layers will influence perovskite lasing actions. Especially, the appearance of perovskite/transport layer interfaces, which are absent in optically pumped perovskite lasers, is usually considered a detrimental factor to light emission in perovskite lightemitting diodes due to the transport layer induced photocarrier quenching. [10] Moving from light-emitting diodes toward lasers, there were reports realizing high injection current densities (>1 kA cm −2 ) by carefully integrating the perovskite films into electrical device structures; yet lasing actions have not been realized. [9,11,12] This might be caused by the reduced optical gain properties or even higher lasing threshold of perovskites when the perovskite films are sandwiched by transport layers in electrical devices. Hence, it is very important to reveal the effects of transport layers on the perovskite lasing threshold and develop excellent transport layers that will maintain the optical gain properties or even reduce the lasing threshold of perovskites when integrating them with perovskite films.
We focus our study on the benchmark methylammonium lead bromide (MAPbBr 3 ) perovskite film because of its low amplified spontaneous emission (ASE) threshold and intensive previous investigations on the optically pumped lasing actions. [13][14][15][16] Here, the ASE is realized on the condition that the population of excited states exceeds the population of ground states (so-called population inversion condition). Since this is the same required condition for the laser emission, the ASE threshold provides a comparable indication for the laser threshold.
In the present work, hole-and electron-transport layers are introduced on top of perovskite thin films and the resulting ASE thresholds are characterized to study their effects on perovskite lasing actions. With introducing an additional hole-transport layer poly(triaryl amine) (PTAA), we surprisingly find that the ASE threshold is reduced by 22.9% and the ASE intensity is improved by 18.8%. These changes cannot be explained by the possible defect passivation effects induced by the additional PTAA layer as PTAA reduces the spontaneous emission of MAPbBr 3 . Further transient photoluminescence (PL) and transient absorption (TA) measurements suggest that the threshold in this case is determined by the hot-carrier cooling process. Specifically, the relatively slow hot-hole cooling process [17][18][19] within the bare MAPbBr 3 film inevitably enlarges the ASE threshold due to increased Auger loss. By extracting the holes with PTAA, the cooling process is facilitated, which largely reduces the Auger loss and thus reduces the ASE threshold. In addition, we also attempt to directly avoid excess hot-hole generation in perovskites by switching the pumping wavelength from 400 to 500 nm, and the resulting ASE threshold is further reduced from 25.7 to 7.2 μJ cm −2 providing another direct evidence to support our argument of the hot-hole effects on the ASE threshold of perovskites. Our work provides practical design strategies toward the realization of electrically pumped perovskite lasers.

Results and Discussion
We first characterize the gain properties of MAPbBr 3 films without transport layers using a femtosecond pulsed laser source (with a central wavelength of 343 nm). The scanning electron microscopy (SEM) image (Figure 1a) indicates that the perovskite film is compact and smooth without pinholes. As shown in Figure 1b, PL spectra under different pump fluences are compared. At very low fluence (14.0 μJ cm −2 ), a relatively broad PL spectrum is observed with a peak at 538 nm and full width at half maximum (FWHM) of 24 nm. When the pump fluence increases, an additional narrow peak at ≈551 nm arises. This additional peak dominates the PL spectrum with further increasing pump fluence, which is a typical signature of ASE. Here, the broad PL at 538 nm corresponds to the spontaneous emission while the additional narrow peak at 551 nm corresponds to the stimulated emission (SE). Figure 1c compares the integrated PL intensity (spontaneous emission + SE) and the FWHM at different pump fluences. When the FWHM starts to increase (at ≈64.2 μJ cm −2 ), the logarithmic slope of PL intensity versus the fluence starts to increase. It clearly reveals the optical gain property of MAPbBr 3 , where light amplification occurs. We would like to point out that the threshold is determined at the point where both integrated PL intensity (optical gain starts) and FWHM (sharp ASE peak appears) change obviously. For most lasing materials, the ASE peak and spontaneous emission peak are close to each other, resulting in the FWHM narrowing at the threshold. [20] However, in our present case, the ASE peak and spontaneous emission peak are spectrally separated, i.e., the ASE peak locates too far away from the spontaneous emission peak (Figure 1b). When the ASE peak appears initially, the whole spectrum broadens, leading to increased FWHM. With continuously increased pumping fluence, the ASE peak becomes dominant and FWHM drops rapidly. Therefore, the ASE threshold is determined to be 64.2 μJ cm −2 for the MAPbBr 3 -only sample. When SE dominates the PL, FWHM decreases sharply ( Figure 1c) together with rapidly increasing PL intensity.
To study the effect of transport layers on the ASE, we select two typical charge-transport layers that show nearly no absorption in the MAPbBr 3 emission range, i.e., PTAA (a hole-transport material) and 4,7-diphenyl-1,10-phenanthroline (Bphen) (an electrontransport material) with the energy diagram shown in Figure  S1 (Supporting Information). The PTAA or Bphen dissolved in chlorobenzene separately is spin-coated on top of the thermally annealed perovskite layer, which means the morphologies of the underneath perovskite layers will not be altered since the perovskites are not soluble in chlorobenzene. This is verified by the same SEM images (Figure 1a To understand the underlying mechanism of the transport layers influenced ASE behaviors of the perovskite films, a series of studies are performed. First, we exclude the possibility of defect passivation effects induced by the additional layers that affect the ASE thresholds and intensities. [21,22] The PL spectra and transient PL of all three samples: MAPbBr 3 film, MAPbBr 3 /PTAA film, and MAPbBr 3 /Bphen film are compared. At relatively low pump fluence (36.0 μJ cm −2 that is below the ASE thresholds for all three samples under the pumping wavelength of 343 nm), only spontaneous emissions exist. As shown in Figure 2a, the PL peaks are the same, indicating that the crystallization of the perovskite film is not affected by coating additional PTAA or Bphen. Interestingly, the PL intensity varies with different transport layer treatments. For MAPbBr 3 /PTAA film, PL intensity significantly drops as compared with the MAPbBr 3 film, indicating an efficient hole extraction process after introducing the additional PTAA layer through the energy-favorable band alignment of MAPbBr 3 and PTAA (see Figure S1, Supporting Information). Such extraction is further verified by transient PL results in Figure 2b, where PL lifetime decreases from 7.8 ns (MAPbBr 3 /PTAA film) to 6.4 ns (MAPbBr 3 film). By contrast, the MAPbBr 3 /Bphen film shows relatively enhanced PL intensity as compared with the MAPbBr 3 film and a nearly unchanged PL lifetime (7.7 ns, see Figure S5, Supporting Information). The unchanged PL lifetime implies the absence of electron extraction, agreeing well with the energy-unfavorable band alignment of MAPbBr 3 and Bphen (see Figure S1, Supporting Information). Regarding the fact that MAPbBr 3 /Bphen sample shows higher PL intensity, but unchanged PL lifetime as compared with the MAPbBr 3 sample, we attribute this to the change of light outcoupling. Perovskite has a much higher refractive index than organic materials, leading to a much lower light outcoupling efficiency. [23] With the insertion of an additional Bphen layer (the refractive index of which is in between that of the perovskite and air), more light can be outcoupled, resulting in enhanced PL intensity. Since outcoupling change does not influence the carrier dynamics, PL lifetime remains unchanged. Here, the shortened PL lifetime of the perovskites after introducing the PTAA layer suggests that defect passivation is not a possible mechanism responsible for the ASE threshold reduction. In addition, considering that all three samples contain the same perovskite layer, and the pumping light source is always excited from the glass side, the absorptions for the perovskite layer are the same. Hence, lower PL intensity at the same pump fluence indicates lower PLQY. Here, the PL comparison results at low pump fluence reveal an interesting trend that the sample with a lower PLQY value shows a lower ASE threshold. Thereby, the defect passivation mechanism is further excluded to explain such an interesting phenomenon (enhancing ASE emission while suppressing spontaneous emission) by introducing an additional PTAA layer.
We also exclude the possibility of the light out-coupling effects on the reduced ASE threshold and enhanced ASE intensities upon the addition of the PTAA layer. Figure 2c shows the PL spectra at a high pump fluence of 705.4 μJ cm −2 . Since the fluence is much higher than the thresholds, the ASE spectra can reveal the gain profiles. Specifically, the ASE intensity of the MAPbBr 3 /PTAA film exhibits a much stronger signal as compared with MAPbBr 3 film, which indicates a higher gain value and is also consistent with the lower ASE threshold. Contrarily, the MAPbBr 3 /Bphen film shows a weaker ASE signal and a higher ASE threshold ( Figure S4, Supporting Information). Combining the fact that the PL lifetimes of the MAPbBr 3 film and the MAPbBr 3 /Bphen film are nearly identical (see Figure  S5, Supporting Information), it is possible that the Bphen layer only affects the light out-coupling of perovskites rather than the carrier dynamics. Since PTAA shares a similar refractive index (≈1.7) as Bphen, [24,25] similar light out-coupling effects that reduce the gain property (SE intensity) can be expected. Thus, the light out-coupling effect of the transport layers is also excluded to explain the simultaneously enhanced ASE emission and suppressed spontaneous emission after combining an additional PTAA layer.
We then explore the carrier dynamics of the samples at high pumping fluence to gain more insight into the ASE behaviors of perovskite films after introducing transport layers. The increment of ASE intensity in the MAPbBr 3 /PTAA film (Figure 2c) indicates that the optical gain is largely enhanced with the PTAA treatment. This enlarged gain is further verified by a largely shortened PL lifetime (from 6.7 ns of the MAPbBr 3 film to 1.2 ns of the MAPbBr 3 /PTAA film) at a pump fluence of 705.4 μJ cm −2 . Here, the short PL lifetime indicates a fast recombination rate via the SE process is dominated after introducing the PTAA layer on top of the perovskite film. To confirm our analysis that the largely decreased PL lifetime at high pump fluence is caused by the enhanced SE process, we introduced TA measurement (Figure 3a and b). As reported, the temporal photobleaching signal reveals the decay of carrier density directly. [17,26] In our experiment, photobleaching signals at 527 nm were tracked (see Figure 3e). Consistently, similar decay curves are observed for the MAPbBr 3 film and the MAPbBr 3 /Bphen film (see Figure S5b, Supporting Information), further verifying our hypothesis that Bphen does not change the carrier dynamics of the perovskite layer under the high pump fluence. However, with introducing the PTAA layer, the decay of the photobleaching signal is significantly facilitated, implying a superfast consumption of photoexcited charge carriers. This phenomenon, together with the enhanced SE intensity and the largely reduced PL lifetime, clearly proves that the additional PTAA layer is beneficial to the gain property of the perovskite layer.
To figure out the underlying mechanism for the improved gain property enabled by the hole-transport layer PTAA, the hot-carrier cooling process is analyzed by tracking the time evolution of the photobleaching peak position. As shown in Figure 3b and Figure S6b (Supporting Information), the photobleaching peak redshifts with time for all three samples, which is a typical feature of the hot-carrier cooling process among perovskites. By using the exponential fitting, the hot carrier cooling lifetime for three samples are 0.49 ps (the MAPbBr 3 film), 0.42 ps (the MAPbBr 3 /PTAA film), and 0.49 ps (the MAPbBr 3 /Bphen film), respectively. Thus, the PTAA layer is proven beneficial to facilitate the hot-carrier cooling process. Considering our observation in Figure 2a and b that PTAA shows good hole-extracting ability, it implies that the hole extraction of PTAA can facilitate the hot-hole cooling process by suppressing the hot-phonon bottleneck effect.
We analyze the influence of the hot-hole-induced Auger loss on net gain coefficient G, to understand its effect on the ASE threshold. As defined, G = G 0 -L, where G 0 is positively related to the accumulated photoexcited carriers at population inversion condition and L refers to the overall processes that compete with the SE process. When Auger loss increases (i.e., L increases), G 0 should also increase to make G = 0 (the threshold condition). Thus, for higher hot-hole-induced Auger loss, higher G 0 is required for reaching the threshold condition, meaning that a larger accumulated carrier density is required (corresponding to higher threshold pump fluence). Here, the facilitated hot-carrier cooling largely suppresses the Auger loss, leading to a lowered ASE threshold. Since perovskites are intrinsically ionic materials where large polarons are formed initially with high-energy photon excitation. [27,28] As schematically shown in Figure 4a, the existence of the large polarons distorts the lattice structure by Coulomb force, which polarizes the perovskite locally around these polarons. Thus, ionic vibration can be enhanced so that the phonon modes increase drastically, which increases the rate of phonon-assisted Auger recombination, a serious loss mechanism against lasing actions at high carrier density. By extracting the holes in perovskites, hot-hole relaxation will be facilitated, leading to reduced Auger loss, and thus favoring better optical gain property and enhanced ASE behavior as schematically illustrated in the left and middle panels of Figure 4c.
Besides reducing Auger recombination loss, the extraction of holes from perovskites can also benefit lasing actions by enhancing the coherent interaction between excited states. [13] Since the perovskite polycrystalline film is a p-type semiconductor with excess holes, [29] spontaneous emission intensity at low excitation intensity can be enhanced due to increased electron-hole-pair formation probability. However, at high excitation intensity, the excess holes with positive charges break the coherent interactions between electron-hole pairs, detrimental to the lasing action. By extracting these excess holes in perovskite film with PTAA, both spontaneous emission reduction at low excitation intensity and ASE enhancement at high excitation intensity can be expected, which is also consistent with our results in Figure 1.
To further validate our hypothesis, we design an independent measurement, where we compare the ASE thresholds of the bare perovskite sample by directly tuning the amount of initial hot carriers. By increasing the pump wavelength, i.e., decreasing the excitation photon energy, the hot carriers will be largely reduced owing to less energy absorbed. [30] When the pump wavelength is tuned to be 400 nm, the integrated PL intensity versus pump fluence curve in Figure 4b exhibits a largely reduced ASE threshold of 25.7 μJ cm −2 that is much lower than the threshold of 64.2 μJ cm −2 when using 343 nm as excitation (Figure 1). By fur-ther increasing the pump wavelength to 500 nm (see the right panel in Figure 4c), i.e., much fewer hot carriers are generated, the ASE threshold is further reduced to 7.2 μJ cm −2 (Figure 4b).
According to the excitation wavelength-dependent PL lifetime results (375 and 405 nm in Figure S7, Supporting Information), the trend that MAPbBr 3 /PTAA shows the shortest PL lifetime remains at different excitation wavelengths. This further verifies that the hole extraction ability of PTAA is regardless of the excitation wavelength. Hence, this independent measurement further proves that hot carriers are detrimental to the optical gain process in metal halide perovskites. Our recent discovery [31] indicates that in both bromide-and iodine-based perovskites, ASE thresholds reduce with decreasing excitation photon energy. It reveals that such a phenomenon is universal. In the present work, we further reveal that Auger recombination loss suppression is the origin of MAPbBr 3 perovskite system; in addition, we find that the hole-transport layer PTAA can also contribute to this process by extracting excess holes. Furthermore, we also compare the ASE threshold between MAPbBr 3 and MAPbBr 3 /PTAA in Figure S8 (Supporting Information) with 508 nm excitation. It shows that these two samples share similar threshold power (≈20 μW). We attribute it to the largely reduced hot-hole generation owing to lower energy excitation. Thus, the Auger loss is not so obvious and the effect of hole extraction to reduce Auger loss is not obvious, resulting in the nearly unchanged ASE threshold.

Conclusion
We have found that the slow hot-hole cooling process is detrimental to the perovskite lasing actions where the Auger loss will increase. By introducing the hole-transport layer PTAA on top, the gain property of the perovskite layer can be largely improved, as shown by reduced ASE threshold and enhanced gain profile. Such improvement is rationalized by linking the facilitated hot-carrier cooling process and efficient hole extraction. Furthermore, the relationship between hot carriers and ASE thresholds is also evidenced by directly switching the pump wavelengths. Our results suggest that the judicious selection/design of transport layers can be beneficial for the perovskite lasing actions, eliminating the community's common concern about their detrimental impact. Such benefit is induced by the suppression of the hot-hole-induced Auger recombination loss by hole extraction, instead of the conventional reported defect passivation mechanism. Our work thus provides a practical strategy to reduce the threshold of perovskite lasing action, which paves the way for the future design of electrically pumped perovskite lasers.
Preparation of the Perovskite Films: The precursor solution of MAPbBr 3 was prepared by dissolving PbAc 2 •3H 2 O (0.3552 g) with MABr (0.3136 g) in DMSO (1 mL). The precursor solutions of the transport layers were prepared by dissolving PTAA or Bphen in CB with a concentration of 3.0 mg mL −1 . All the perovskite films were fabricated on glass substrates, which were precleaned by the ultrasonic treatment, blown dry with nitrogen gas and treated with ultraviolet-ozone. For the MAPbBr3-only sample, 50 μL of MAPbBr 3 precursor solution was spin-coated on the glass substrates at 500 rpm. for 7 s and then 3000 rpm. for 60 s, and treated with thermal annealing at 60°C for 30 min. For the MAPbBr 3 /PTAA and MAPbBr 3 /Bphen samples, the precursor solutions of corresponding transport materials were spin-coated on top of the thermally annealed MAPbBr 3 film at the speed of 4000 rpm for 60 s, and then annealed at 60°C for 10 min. All the films were encapsulated by a thin glass and epoxy resin in the glovebox for experimental measurements.
Characterization: All the PL spectra were collected using femtosecond laser systems. Specifically, for 343 nm generation, a harmonic generator (Ultrafast Systems LLC, third harmonic) pumped by a Pharos laser (Light Conversion, 1 kHz, 1030 nm, 290 fs) was used. For 400 and 500 nm generation, a continuous-wave green laser (peaked at 532 nm) initiated by Verdi V-10 (Coherent) was transformed into an ultrashort pulsed laser by Mira model 900 (Coherent) and amplified by Legend Elite (Coherent), and further modulated by an amplifier (OPA, from OPerA-Solo (Coherent)). The PL signals were measured with a Flouro Log III spectrometer. TA spectra were collected with a Helios Fire spectrometer (Ultrafast Systems LLC) and the probe beam (broad light emission) was generated by CaF 2 crystal excited by a 1030 nm fs primary beam.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.