Advanced Photonics, Volume. 7, Issue 5, 056006(2025)

Volatile ammonium-driven perovskite phase reconstruction for high-performance quasi-CW lasing

Xinyang Wang1, Guochao Lu1, Qiuting Cai1, Jing Li2、*, Haoran Zhang1, Zaishang Long1, Meiyi Zhu3, Yun Gao1, Qingli Cao1, Hanyan Huang1, Xingliang Dai1,3,4,5、*, Zhizhen Ye1,3,4,5, and Haiping He1,3,4,5、*
Author Affiliations
  • 1Zhejiang University, School of Materials Science and Engineering, State Key Laboratory of Silicon and Advanced Semiconductor Materials, Hangzhou, China
  • 2Zhejiang University of Technology, Science and Education Integration College of Energy and Carbon Neutralization, College of Materials Science and Engineering, Zhejiang Provincial Key Laboratory of Clean Energy Conversion and Utilization, Hangzhou, China
  • 3Zhejiang University, Institute of Wenzhou, Wenzhou Key Laboratory of Novel Optoelectronic and Nano Materials, Zhejiang Provincial Engineering Research Center, Wenzhou, China
  • 4Shanxi‐Zheda Institute of Advanced Materials and Chemical Engineering, Taiyuan, China
  • 5Wenzhou Xinxin Taijing Technology Co., Wenzhou, China
  • show less

    All-inorganic CsPbBr3 perovskite polycrystalline films, renowned for their remarkable optoelectronic properties, solution processability, and enhanced stability over organic–inorganic counterparts, are emerging as next-generation gain media for high-performance lasers. However, due to the limited understanding of how to realize population inversion under slow carrier injection, and a lack of convenient strategies to suppress Auger recombination while retaining low optical loss, achieving high-performance quasi-continuous-wave (quasi-CW) or CW lasing based on CsPbBr3 films at ambient temperature is still challenging. We devised a phase reconstruction strategy employing volatile ammonium, which achieves substantial suppression of Auger recombination through elimination of low-dimensional phase impurities and remains low optical loss via precisely controlled film crystallization dynamics. Importantly, this strategy emphasizes the critical role of Auger recombination suppression for high-performance lasing under slower carrier injection. Ultimately, an ultralow amplified spontaneous emission threshold of 9.6 μJ cm - 2 was achieved under quasi-continuous nanosecond-pulsed excitation, which facilitated the realization of a single-mode vertical-cavity surface-emitting laser with a threshold of 17.3 μJ cm - 2 and a quality factor of 3850 under quasi-CW pumping. We represent the exceptional performance of quasi-CW lasing to date, offering valuable insights for future advancements in high-performance CW lasing and even electrically driven lasers.

    Keywords

    1 Introduction

    Metal halide perovskite semiconductors have garnered widespread attention as coherent light sources owing to their high absorption coefficient, balanced long-range carrier diffusion, and high defect tolerance.19 In particular, all-inorganic CsPbBr3 perovskite films, prepared via a one-step spin-coating method, are highly favored as next-generation gain media due to their better integrability, simplified fabrication processes, and enhanced stability compared with their organic–inorganic counterparts.1012 It should be emphasized that an excellent film gain medium necessitates a compact, smooth surface to reduce optical loss for good optical gain performance.1315 Unfortunately, the poor solubility of cesium bromide (CsBr) in precursor results in low coverage and a highly rough surface of the pristine CsPbBr3 films, severely impeding the occurrence of amplified spontaneous emission (ASE) and lasing.11,12,16 The incorporation of bulky organic ammonium cations to form quasi-two-dimensional (quasi-2D) phases addresses this issue effectively. Bulky organic ammonium cations, such as PEA+ (C8H9NH3+) and BA+ (C4H9NH3+), contribute to the confinement of perovskite grains and improve the surface morphology of the films, which have been demonstrated to be viable for constructing low-threshold perovskite lasers.1719

    However, Auger recombination poses a considerable challenge to the development of quasi-2D perovskite lasers under quasi-continuous-wave (quasi-CW) or CW excitation.18,20 Current research has successfully enhanced the morphology of CsPbBr3 films and enabled lasing under femtosecond- or subnanosecond-pulsed excitation,10,11,13 but with limited emphasis on Auger recombination, especially under the condition of slower carrier injection. Therefore, it is essential to comprehend the influence of Auger recombination on lasing performance. According to the Bernard–Duraffourg condition, the carrier concentration in perovskites must reach at least 1017  cm3 at room temperature to realize stimulated emission.21,22 At high excitation fluence, Auger recombination and stimulated emission compete as rapid carrier decay pathways; however, the carrier density required for Auger recombination is lower than that for stimulated emission.23 As a result, Auger recombination usually dominates the carrier recombination dynamics under such high carrier density, which directly hinders the accumulation of carrier concentration and thus limits the realization of population inversion.24 Notably, the introduction of bulky organic ammonium cations often induces quasi-2D phases, and the strong dielectric and quantum confinement in quasi-2D systems further increases the exciton collision probabilities and enhances the likelihood of Auger recombination.25,26 In addition, the naturally formed multiquantum-well structures are typically irregular and nonuniform, causing a localization of charge carriers in undesired areas and thus exacerbating nonradiative Auger recombination.18,27 Although ultrafast femtosecond-pulsed excitation can effectively mitigate the impact of Auger recombination in quasi-2D systems, achieving high-performance lasing under extended-pulse-width excitation remains a formidable challenge due to the slower carrier injection.17 Therefore, to achieve high-performance lasing under nanosecond-pulsed excitation or even CW excitation, suppressing Auger recombination is imperative.

    In this study, we developed a gain medium with both low optical loss and suppressed Auger recombination by employing a volatile ammonium-driven phase reconstruction strategy, enabling the realization of low-threshold single-mode lasing under quasi-CW excitation. Unlike the commonly used 2-phenylethylamine hydrobromide (PEABr), the addition of volatile 2-phenylethylammonium acetate (PEAAc) drives phase reconstruction during the annealing of polycrystalline films, substantially improving film quality and achieving gain media with low optical loss as well as reduced Auger rates. Importantly, this study underscores that the suppression of Auger recombination is essential for achieving optimal lasing performance under extended pulse-width excitation. Benefiting from the effective suppression of Auger recombination, the PEAAc-modified CsPbBr3 films exhibit an ASE threshold of 9.6  μJcm2, which is ultra-low for polycrystalline films under quasi-continuous nanosecond-pulsed excitation. Building on this, we successfully constructed a vertical-cavity surface-emitting laser (VCSEL) and obtained high-performance single-mode lasing with a threshold of 17.3  μJcm2 and a quality factor of 3850 under nanosecond-pulsed excitation, which represents the best quasi-CW lasing performance to date. This work highlights the critical role of Auger recombination in achieving lasing under extended pulse-width excitation, providing valuable insights for the exploration of CW and even electrically driven lasers.

    2 Results and Discussion

    2.1 Mechanism of Volatile Ammonium-Driven Phase Reconstruction

    The pristine CsPbBr3 precursor was prepared by dissolving equimolar amounts of CsBr and lead bromide (PbBr2) in dimethyl sulfoxide (DMSO). To enhance the morphology of the resulting films, two types of ammonium salts, PEABr and PEAAc, were incorporated into the CsPbBr3 precursors, named w/PEABr and w/PEAAc, respectively. The perovskite films were fabricated utilizing a one-step spin-coating technique, followed by an annealing process at 100°C for 10 min. More details are shown in the Section of “ASE and lasing experiments” in the Supplementary Material. Unlike PEABr, a commonly used ammonium salt with a high decomposition temperature of 200°C, PEAAc demonstrates notable volatility, as evidenced by weight loss starting around 70°C (Fig. S1 in the Supplementary Material). The addition of volatile PEAAc offers a new approach to controlling the crystallization of perovskite films, as depicted in Fig. 1. The pristine CsPbBr3 film, with poor coverage and rough surface, suffers from considerable optical losses, making it almost unsuitable as a gain medium.28 The incorporation of the commonly used nonvolatile PEABr induces the formation of quasi-2D perovskites to eliminate pinholes effectively, resulting in the formation of high-quality films with dense and smooth surfaces.20 However, it concurrently results in strong quantum and dielectric confinement effects,25 which exacerbate Auger recombination and diminish optical gain. Notably, the introduction of volatile PEAAc effectively addresses the aforementioned deficiencies simultaneously. The PEA+ cations in PEAAc are strongly coordinated with inorganic compounds in the precursor, resulting in the formation of numerous homogeneous colloids that serve as growth sites. During subsequent thermal annealing, the intrinsic volatility of PEAAc facilitates its separation from the perovskite, thereby liberating grain growth constraints. Ultimately, the w/PEAAc film exhibits high crystallization quality with minimal quantum and dielectric confinement effects, holding the promise for achieving low optical loss and reduced Auger recombination rates, showing potential as a gain medium for high-quality lasers.

    Volatile ammonium-driven phase reconstruction.

    Figure 1.Volatile ammonium-driven phase reconstruction.

    Through comprehensive experimental characterization and in situ observations, we validated the chemical processes and outcomes depicted in the schematic. Dynamic light scattering was first employed to examine the behavior of colloids in the precursors. According to the Gibbs free energy (includes surface units and volume units) curves [Fig. S2(a) in the Supplementary Material], crystal nuclei can grow spontaneously only if the radius exceeds the critical size r*; otherwise, it will disintegrate.29,30 In the pristine CsPbBr3 precursor, the large colloids that can exist stably only account for a small proportion and show a wide distribution, which is not conducive to the uniform growth of crystal nuclei to form a dense and smooth film. After introducing PEABr or PEAAc into the solution, a large number of colloids, predominantly around 300 nm in size, can be observed [Fig. S2(b) in the Supplementary Material]. The abundant and stable colloids with narrow size distributions can act as growth sites, promoting uniform surface coverage.

    To further understand the role of PEAAc in phase reconstruction, in situ photoluminescence (PL) and time-dependent ultraviolet–visible spectroscopy (UV–vis) were utilized to monitor the reaction processes of PEAAc and PEABr with inorganic compounds during the entire annealing process [Figs. S3 and S4 in the Supplementary Material]. Notably, the PL spectra were normalized at each time point to highlight the evolution in peak shape. For comparative analysis, the spectra for the initial 90 s are specifically extracted to emphasize the variations. In the w/PEAAc films, a narrow PL signal centered at 407 nm and a broad PL signal spanning 430 to 515 nm appear at the beginning of the annealing procedure, as shown in Fig. 2(a), which can be attributed to the n=1 phase and n2 phases, respectively.31,32 As the annealing progresses, the higher-energy emission decreases and eventually disappears, whereas the lower-energy PL peak narrows and experiences a redshift. After 90  s of annealing, only a sharp peak remains at 523 nm, corresponding to three-dimensional (3D) phases.33 The evolution of the PL peaks indicates that the w/PEAAc films transition from quasi-2D perovskites to 3D perovskites during thermal annealing, the process of phase reconstruction, driven by the desorption of PEAAc from the inorganic compounds. By contrast, both w/PEABr films and pristine CsPbBr3 films consistently exhibit an unaltered PL peak throughout the whole annealing process [Figs. 2(b) and 2(c)]. Time-dependent absorption spectra further confirm these observations. As shown in Fig. 2(d), the initial state of w/PEAAc film exhibits several excitonic absorption peaks at 403, 434, and 466 nm, corresponding to the domains with n=1, 2, and 3, respectively.17,31,32 These peaks degrade and fade away as the annealing time increases, whereas the absorption edge at 515 nm evolves into a prominent excitonic absorption peak, indicating the occurrence of phase reconstruction. However, the w/PEABr films and the CsPbBr3 films exhibit substantially coincident absorption curves throughout the whole annealing process [Figs. 2(e) and 2(f)]. Notably, the w/PEABr films exhibit multiple absorption peaks attributed to different low-dimensional phases,34 whereas the CsPbBr3 films exhibit typical characteristics of three-dimensional phase absorption.17,35 Moreover, X-ray photoelectron spectroscopy (XPS) analysis reveals that the Pb 4f core-level spectra in the w/PEABr films after annealing shift toward higher binding energy compared with the pristine CsPbBr3 films, verifying the strong coordination between PEA+ and Pb2+. On the contrary, no significant shift is observed in the w/PEAAc films compared with the pristine CsPbBr3 films, further demonstrating the desorption of PEAAc after annealing [Fig. 2(g)].

    Mechanism illustration of the phase reconstruction process. In situ PL spectra of w/PEAAc (a), w/PEABr (b), and pristine CsPbBr3 films (c), as well as time-dependent UV spectra of w/PEAAc (d), w/PEABr (e), and pristine CsPbBr3 films (f), focusing on the initial 90 s of thermal annealing. (g) XPS core-level spectra of Pb 4f. H1 NMR spectra of w/PEAAc and w/PEABr before annealing (h) and after annealing (i).

    Figure 2.Mechanism illustration of the phase reconstruction process. In situ PL spectra of w/PEAAc (a), w/PEABr (b), and pristine CsPbBr3 films (c), as well as time-dependent UV spectra of w/PEAAc (d), w/PEABr (e), and pristine CsPbBr3 films (f), focusing on the initial 90 s of thermal annealing. (g) XPS core-level spectra of Pb 4f. H1 NMR spectra of w/PEAAc and w/PEABr before annealing (h) and after annealing (i).

    To determine whether the separable PEAAc remains in the films after annealing, Fourier-transform infrared (FTIR) spectra and H1 nuclear magnetic resonance (H1-NMR) spectra were employed to address the issue. The FTIR spectra reveal that the w/PEABr films show intense signals of N─H and C─H vibration peaks after thermal annealing,30 whereas the w/PEAAc films and the pristine CsPbBr3 films do not [Fig. S5(a) in the Supplementary Material]. Further analysis of the FTIR spectra of the w/PEAAc films at different annealing times revealed a gradual reduction in the intensity of the N─H, C═O, and C─H vibration peaks [Fig. S5(b) in the Supplementary Material].30,36 These peaks, originating from PEAAc, finally vanished, evidencing the absence of groups such as ─C═O─, ─C─H─, and ─N─H─ in the annealed films. The H1 NMR spectra further confirm the separability of PEAAc. The fresh w/PEAAc films, as well as those containing PEABr, display multiple distinct H1 NMR peaks corresponding to hydrogen atoms in various chemical environments within the Ac groups and PEA+ cations [Fig. 2(h)], indicating the presence of ammonium salts in these films before the annealing process. After annealing, however, the w/PEAAc films no longer display these signals in the H1 NMR spectra, whereas the w/PEABr films maintain their H1 peaks [Fig. 2(i)]. In addition, the XPS spectra of the annealed w/PEAAc films lack the N 1s peak (Fig. S6 in the Supplementary Material). These results collectively confirm that PEAAc completely escapes from the films after annealing, leaving no residual components.

    To further explore grain evolution during thermal annealing, scanning electron microscopy characterization was conducted on both fresh and annealed films. The grain sizes of both the pristine CsPbBr3 films and the w/PEABr films remain relatively unchanged throughout the annealing process (Fig. S7 in the Supplementary Material). Specifically, although the incorporation of PEABr enhances film continuity, it also restricts grain growth, resulting in smaller grains. Interestingly, a distinct increase in grain size is observed in the w/PEAAc films post-annealing (Fig. S8 in the Supplementary Material). The results of X-ray diffraction reveal significantly enhanced diffraction peaks in the w/PEAAc films compared with both pristine CsPbBr3 and w/PEABr films after annealing, confirming that the addition of PEAAc can induce better crystallinity (Fig. S9 in the Supplementary Material). Concurrently, atomic force microscopy images demonstrate that the films incorporating organic cations exhibit a more uniform and compact surface morphology compared with the pristine CsPbBr3 films (Fig. S10 in the Supplementary Material). The low optical loss of the w/PEAAc films is further corroborated by the PL loss coefficient, which was determined by measuring the PL intensity as a function of distance (Fig. S11 in the Supplementary Material).

    2.2 Recombination Dynamics of Perovskite Films

    We measured the PL spectra under various excitation densities to determine the dominant emission mechanism in our samples, and the results are shown in Fig. S12 in the Supplementary Material. The PL integrals of both w/PEAAc and w/PEABr films show power-law dependence on the excitation power, with exponents of 1.32 and 1.12, respectively. In direct bandgap semiconductors and under nonresonant excitation conditions, the integrated PL intensity (IPL) is a power-law function of the excitation density,37IPLIexk,with k<1 for free-to-bound recombination and donor–acceptor pair, 1<k<2 for the recombination of excitons (including free excitons and bound excitons), and k=2 for free carrier recombination. The model was further refined by Shibata et al.,38 who provided an analytical formula confirming that 1<k<2, even for free excitons. Our k value agrees well with the emission from excitonic states. Notably, the deviation from the power-law relationship at high excitation densities can be attributed to nonradiative Auger recombination.17 The significant differences in both the inflection points and the degrees of deviation between the two samples suggest distinct carrier dynamics under strong excitation conditions for each sample.

    To investigate the role of PEAAc-driven phase reconstruction in reducing quantum and dielectric confinement, the exciton binding energy (Eb) of the w/PEAAc and w/PEABr films was extracted using temperature-dependent PL measurements.39 As shown in Figs. 3(a) and 3(b), the decrease in PL intensity with increasing temperature can be attributed to luminescence quenching resulting from the dissociation of excitons. The Eb values of the w/PEAAc and w/PEABr films are estimated to be 66.5 and 88.7 meV, respectively, both of which exceed the thermal energy at room temperature (26  meV), indicating the feature of exciton recombination (a monomolecular process) in both films.40 The high Eb of our samples, combined with the k values and distinct exciton absorption peaks observed in Figs. S4(a) and S4(b) in the Supplementary Material, effectively demonstrate the exciton recombination mechanism in w/PEAAc and w/PEABr films. In principle, a high Eb indicates a strong confinement effect, often associated with rapid Auger recombination.25,4144 The extracted Eb confirms that the addition of volatile PEAAc contributes to weakening the confinement and thus should be conducive to suppressing Auger recombination. To further substantiate this finding, the relative photoluminescence quantum yields (relative PLQY) of the films, derived from the ratio of integrated light output to light input, were tracked as a function of excitation power to qualitatively assess Auger recombination, as depicted in Fig. 3(c). The method for calculating carrier concentration is detailed in Fig. S13 in the Supplementary Material. At low excitation densities, the PLQY of both w/PEAAc films and w/PEABr films rises due to the occupation of defect states.45 As the carrier concentration increases, Auger recombination gradually dominates the carrier recombination dynamics in the w/PEABr films, leading to a decline in relative PLQY.46 By contrast, the w/PEAAc films exhibit a higher excitation threshold at which PLQY reaches saturation. Notably, no additional radiative recombination processes, other than exciton recombination, were observed in the test [Figs. S13(a) and S13(b) in the Supplementary Material], and film degradation as a cause for the decrease in PLQY has been ruled out by forward and inverse measurements [Figs. S13(c) and S13(d) in the Supplementary Material]. These results indicate that Auger recombination is significantly suppressed in the w/PEAAc films compared with the w/PEABr films.

    Recombination dynamics for w/PEAAc films and w/PEABr films. Temperature-dependent PL spectra of w/PEAAc films (a) and w/PEABr films (b). Inset: variation of the integrated intensity with the reciprocal of temperature. The Eb values can be described by the expression of I(T)=I0/(1+aT3/2e−Eb/kT), where T is the temperature and k is the Boltzmann constant. (c) Film’s PLQY as a function of carrier density for w/PEAAc films and w/PEABr films, respectively. Bleaching kinetics with different pump densities for w/PEAAc films (d) and w/PEABr films (e). (f) Auger recombination lifetimes extracted from TA spectra of w/PEAAc and w/PEABr films at ∼4.9 μJ cm−2.

    Figure 3.Recombination dynamics for w/PEAAc films and w/PEABr films. Temperature-dependent PL spectra of w/PEAAc films (a) and w/PEABr films (b). Inset: variation of the integrated intensity with the reciprocal of temperature. The Eb values can be described by the expression of I(T)=I0/(1+aT3/2eEb/kT), where T is the temperature and k is the Boltzmann constant. (c) Film’s PLQY as a function of carrier density for w/PEAAc films and w/PEABr films, respectively. Bleaching kinetics with different pump densities for w/PEAAc films (d) and w/PEABr films (e). (f) Auger recombination lifetimes extracted from TA spectra of w/PEAAc and w/PEABr films at 4.9  μJcm2.

    A comprehensive analysis of carrier recombination dynamics was conducted using femtosecond transient absorption (fs-TA) measurements. The TA spectra of the ammonium-treated films, acquired under 400 nm (200 fs, 100 kHz) laser excitation at an excitation power density of 2.7  μJcm2, are presented in Fig. S14 in the Supplementary Material. Both samples exhibit prominent ground-state bleaching signals near the band edge of the three-dimensional CsPbBr3 phase, which arise from state filling.47 The time evolution of the bleaching signals was further investigated under varying excitation conditions, and the corresponding kinetics were extracted, as illustrated in Figs. 3(d) and 3(e). As the pump fluences increase, rapid decay components on the picosecond time scale become apparent. At an excitation fluence of 4.9  μJcm2, which corresponds to a carrier density of 4.04×1017  cm3 for the w/PEAAc films and 3.18×1017  cm3 for the w/PEABr films, the analysis of the kinetic curves reveals rapid decay component lifetimes of 252 ps for the w/PEAAc films and 113 ps for the w/PEABr films, respectively [Fig. 3(f)]. Given that no stimulated emission occurs at this excitation intensity (Fig. S15 in the Supplementary Material) and that the exciton recombination lifetime exceeds 1 ns, the rapid decay process can be attributed to Auger recombination.12 For the curves in Figs. 3(d) and 3(e), the timing of individual measurements can be adjusted along the time axis to correspond with the next higher carrier density, as shown in Figs. S16(a) and S16(b) in the Supplementary Material. We observed that the TA signal decreases smoothly across several orders of magnitude. From this analysis, we determined that the recombination rate is solely dependent on the excitation density. To estimate the recombination dynamics, we multiplied the TA kinetics by the initial carrier density (n0) to obtain the excitation density [n(t)].48 By differentiating n(t) with respect to time [dndt(t)] and plotting it against n(t), we determined the carrier density-dependent recombination rates, which are presented in Figs. S16(c) and S16(d) in the Supplementary Material. For both types of samples, we found that the recombination rates dndt(t) exhibit two-stage scaling regimes. At low carrier densities (n<5×1016  cm3), the rates are close to linear on the carrier density, whereas at higher carrier densities (n>5×1016  cm3), we found approximate quadratic dependence. Given that exciton recombination predominates near-band-edge emission in both samples, we ascribed the first-order process at low excitation to exciton recombination and trap-assisted recombination, and attributed the second-order process at high excitation to Auger recombination. As no high-order recombination processes were detected, we adopted the general rate equation for the coexistence of mono- and bimolecular charge carrier recombination pathways for fitting,17,45dndt=G+k1n+k2n2,where G is the carrier generation term, n is the exciton density, t is the decay time, and k1 and k2 are the monomolecular recombination (including radiative exciton recombination and trap-assisted recombination) rate and bimolecular Auger recombination rate coefficients, respectively. Note that the contribution of the generation term G can be excluded using ultrashort femtosecond laser pulses.39 As anticipated, we found that k2 of the w/PEAAc films is more than 1 order of magnitude lower than that of the w/PEABr films, reaffirming that the Auger recombination rate has been notably suppressed, as shown in Table 1. We also noticed the decrease of k1 in the w/PEAAc films, which can be attributed to the decline of the exciton recombination constant caused by the reduction of exciton binding energy.49

    • Table 1. Rate coefficients obtained from fitting the fs-TA dynamics of w/PEAAc and w/PEABr films.

      Table 1. Rate coefficients obtained from fitting the fs-TA dynamics of w/PEAAc and w/PEABr films.

      Samplesk1 (s1)k2 (cm3s1)
      w/PEAAc1.44×1075.53×109
      w/PEABr8.90×1073.21×108

    2.3 High-Performance Gain Medium Based on Suppressed Auger Recombination

    Furthermore, the properties of stimulated emission were investigated under femtosecond-pulsed laser pumping (fs-pumping, 400 nm, 200 fs, 1000 Hz). Due to significant optical loss (Figs. S10 and S11 in the Supplementary Material), no ASE characteristics were observed in the spectra of pristine CsPbBr3 films within the pump range of 10.1 to 124.8  μJcm2 (Fig. S17 in the Supplementary Material). For the w/PEAAc films, as shown in Fig. 4(a), a broad spontaneous emission (SE) peak centered at 523 nm with a full width at half maximum (FWHM) of 21 nm is evident at low pump fluence. As the pump fluence exceeds a certain threshold, a sharp and narrow peak emerges on the low-energy side of the SE peak and grows rapidly with increasing pump fluence, indicating the onset of the ASE process.50,51 Time-resolved PL spectra, obtained using a streak camera under excitation fluences both below and above the ASE threshold, are presented in Fig. 4(b). The emission lifetime is dramatically reduced from the nanosecond range to tens of picoseconds, further confirming the occurrence of stimulated emission.19 The minimum ASE threshold for the w/PEAAc films is determined to be 6.7  μJcm2 under fs-pumping, which is approximately half that of the w/PEABr films [13.6  μJcm2, Fig. S18(a) in the Supplementary Material]. The net gain coefficient was evaluated using the variable stripe length (VSL) method,52 fitted by the expression of IASE(L)=IsAG×(eGL1), where IASE(L) is the ASE intensity, G is the net gain coefficient, Is is the pump energy intensity, A is a constant associated with the spontaneous emission cross-sectional area of the material, and L is the stripe length. As shown in Fig. 4(c), the net gain coefficients for the w/PEAAc films are determined to be 293.3, 639.0, and 2048.5  cm1 under 2Pth (threshold power density), 5Pth, and 20Pth excitation, respectively, indicating a strong capacity for light amplification.

    ASE characterizations of the perovskite thin films. (a) Pump fluence-dependent PL spectra of w/PEAAc films under fs-pumping (400 nm, 200 fs, 1 kHz). Inset: ASE integrated intensity and FWHM as a function of pump fluence. (b) Streak camera images showing time-resolved PL emission under the excitation fluence of below (top) and above (bottom) the ASE threshold. (c) Variation of the ASE intensity with stripe length at different pump fluences. (d) Pump fluence-dependent PL spectra of w/PEAAc films under ns-pumping (355 nm, 1.3 ns, 100 Hz). Inset: ASE integrated intensity and FWHM as a function of pump fluence. (e) Evolution of ASE intensity under continuous irradiation of ns-pumping. (f) ASE threshold for w/PEAAc and w/PEABr films under ns-pumping and fs-pumping. (g) Comparison of carrier decay curves extracted from TA spectra and a laser pulse duration of 1 ns. (h) Schematic diagram of rapid Auger recombination hindering carrier accumulation. (i) Tunable ASE over a broad visible light spectrum by halide compositional modulation.

    Figure 4.ASE characterizations of the perovskite thin films. (a) Pump fluence-dependent PL spectra of w/PEAAc films under fs-pumping (400 nm, 200 fs, 1 kHz). Inset: ASE integrated intensity and FWHM as a function of pump fluence. (b) Streak camera images showing time-resolved PL emission under the excitation fluence of below (top) and above (bottom) the ASE threshold. (c) Variation of the ASE intensity with stripe length at different pump fluences. (d) Pump fluence-dependent PL spectra of w/PEAAc films under ns-pumping (355 nm, 1.3 ns, 100 Hz). Inset: ASE integrated intensity and FWHM as a function of pump fluence. (e) Evolution of ASE intensity under continuous irradiation of ns-pumping. (f) ASE threshold for w/PEAAc and w/PEABr films under ns-pumping and fs-pumping. (g) Comparison of carrier decay curves extracted from TA spectra and a laser pulse duration of 1 ns. (h) Schematic diagram of rapid Auger recombination hindering carrier accumulation. (i) Tunable ASE over a broad visible light spectrum by halide compositional modulation.

    Understanding the stimulated emission behavior of perovskites under extended pulse-width excitation conditions is crucial for the development of continuous-wave pumped lasers and, ultimately, electrically pumped lasers. Therefore, we further determined the ASE threshold under nanosecond-pulsed laser pumping (ns-pumping, 355 nm, 1.3 ns, 100 Hz). Notably, the carrier lifetime of the w/PEAAc films, extracted from the TA spectra under excitation of 5.3  μJcm2, is 1.0  ns (Fig. S19 in the Supplementary Material), which is shorter than the pulse duration of ns-pumping (1.3 ns). The carrier lifetime decreases even further as the excitation approaches the ASE threshold. Thus, ns-pumping can be regarded as the quasi-CW pumping condition for the w/PEAAc films.17,53Figure 4(d) illustrates that the w/PEAAc films exhibit a threshold of 9.6  μJcm2 under quasi-continuous ns-pumping, which is ultra-low for perovskite polycrystalline films (Table S1 in the Supplementary Material). Moreover, the ASE intensity of the w/PEAAc films remains stable after continuous irradiation for 5 h at 4Pth of ns-pumping, corresponding to 1.8×106 laser pulses [Fig. 4(e)]. By contrast, the ASE threshold of w/PEABr films significantly increases to 109.3  μJcm2 under ns-pumping [Fig. S18(b) in the Supplementary Material]. It is important to note that switching the excitation from fs-pumping to ns-pumping significantly increases the ASE threshold ratio between the w/PEABr films and the w/PEAAc films, escalating from 2-fold to nearly 10-fold [Fig. 4(f)]. Here, we found that the w/PEABr films remain stable under 2Pth of the ns-pumping condition, suggesting that the elevated ASE threshold is not attributable to film degradation (Fig. S20 in the Supplementary Material). The increase of the ASE threshold may result from pump-induced heating.54 Here, we used an infrared-thermal camera to monitor the surface temperature under different excitation. Under both 10Pth of fs-pumping and 10Pth of ns-pumping, the temperature of the sample remained close to room temperature (24°C), as shown in Fig. S21 in the Supplementary Material, ruling out the possibility. These findings underscore the superior stimulated emission performance of w/PEAAc films under quasi-CW pumping conditions.

    To gain deeper insights into these observations, three critical factors influencing ASE were examined: optical loss, trapping states, and Auger recombination.11 It should be mentioned that the contribution of optical loss can be easily excluded by the approximate roughness and PL loss coefficients of the w/PEAAc and w/PEABr films (Figs. S10 and S11 in the Supplementary Material). As for trapping states, we have conducted the TA spectra of w/PEAAc and w/PEABr films under 0.05  μJcm2 and extracted carrier decay curves in Fig. S22 in the Supplementary Material. We focused on the carrier decay process during the initial 200 ps, which directly competes with stimulated emission. Considering that the contribution of Auger recombination is relatively minor at low excitation, the carrier decay occurring within a few picoseconds in w/PEAAc films primarily results from the rapid trapping process of shallow defects.55 However, w/PEABr films do not exhibit significant carrier attenuation within the initial 200 ps, which can be attributed to the passivation of defects by PEABr and the carrier replenishment from low to large n phases. Based on this, we ruled out the possibility that the improvement of stimulated emission performance by PEAAc originated from defect passivation.

    Therefore, Auger recombination emerges as a decisive role in influencing the gain properties. Under fs-pumping, the carrier injection time (fs) is substantially shorter than the Auger recombination lifetime. Consequently, the accumulation of carriers is minimally affected by Auger recombination. Therefore, under fs-pumping, the difference in the ASE threshold between the w/PEAAc and w/PEABr is primarily governed by the competition between Auger recombination and stimulated emission processes.26 Notably, even though the optical loss coefficient and defect state density of the w/PEABr films are lower than those of the w/PEAAc films, the ASE threshold under femtosecond pumping for the w/PEABr films remains twice as high as that for the w/PEAAc films. This clearly highlights the critical role of Auger recombination in the stimulated emission process. Moreover, the detrimental effects of Auger recombination become more pronounced under pumping conditions with longer laser pulse widths, such as nanosecond pulses or even continuous-wave pumping, where the timescale of carrier injection becomes comparable to or exceeds the Auger recombination lifetime. For instance, Auger recombination significantly affects the decay of carriers on the nanosecond scale under high excitation conditions [Fig. 4(g)]. This results in carrier loss due to Auger recombination during the injection process, which hinders the accumulation of carriers necessary for achieving population inversion and, consequently, limits the occurrence of stimulated emission, as shown in Fig. 4(h). Therefore, suppressing Auger recombination is essential for maintaining optimal gain performance across varying pulse widths of laser excitation. This explains the consistently lower thresholds of the PEAAc films across different pump pulse widths and the more than an order-of-magnitude increase in the threshold for the w/PEABr films when transitioning from fs-pumping to ns-pumping. In addition, the w/PEAAc films without any encapsulation or surface protection exhibit ASE characteristics with a low threshold even after a week of storage in air (Fig. S23 in the Supplementary Material), which indicates the feasibility of the w/PEAAc films as a robust laser medium for long-term operation in air. Furthermore, we achieved tunable stimulated emission through halide compositional modulation, spanning a broad range from deep blue to deep red within the visible light spectrum, as depicted in Fig. 4(i). The average ASE threshold for various wavelengths is below 16  μJcm2 (Figs. S24 and S25 in the Supplementary Material), demonstrating the potential of PEAAc-modified films for application in color-tunable lasers. Furthermore, we confirmed that phenylbutanammonium acetate (PBAAc) and butylammonium acetate (BAAc), as additives, have a similar promoting effect on phase reconstruction, as demonstrated by time-dependent UV spectra and XPS spectra of the films (Figs. S26 and S27 in the Supplementary Material). In addition, ASE characterization under ns-pumping (Fig. S28 in the Supplementary Material) confirmed the universality of the ammonium acetate additive for the fabrication of high-quality gain media.

    2.4 Single-Mode Lasing under Quasi-Continuous-Wave Excitation

    To obtain efficient optical feedback for a high-gain medium, a VCSEL was fabricated by embedding the w/PEAAc film between a pair of distributed Bragg reflector (DBR) mirrors, as shown in Fig. 5(a). A PMMA layer was deposited on the perovskite surface to finely tune the length of the cavity. Each DBR is composed of 11 periods of alternating SiO2 and Ta2O5 layers (the thickness of each SiO2 layer and Ta2O5 layer per period is controlled at 87 and 57 nm, respectively), with their reflection spectra (Fig. S29 in the Supplementary Material) specially designed. The spontaneous emission (red solid line) and lasing (green solid line) peaks are included in the broad wavelength range of high reflectivity (R>99.9%), whereas the excitation wavelength of 355 nm (ns-pumping) and 400 nm (fs-pumping) falls into the high transmission fringe of the DBRs. Figure 5(b) presents the PL spectra under various quasi-continuous ns-pumping fluence levels at room temperature, displaying single-mode quasi-CW lasing centered at 539 nm. The beam output shown in the inset of Fig. 5(b) further confirms the occurrence of quasi-CW lasing.64 Thanks to the high-quality resonator, the single-mode lasing exhibits a narrow FWHM of 0.14 nm, indicating a quality factor (Q factor) of 3850, as shown in the inset of Fig. 5(c). Moreover, the lasing threshold of 17.3  μJcm2 is determined by the integrated intensity evolution as a function of pump fluence [Fig. 5(c)]. Notably, the attainment of the low threshold of 17.3  μJcm2 and the high Q factor of 3850 under ns-pumping in the same perovskite laser is exceptional [Fig. 5(d) and Table S2 in the Supplementary Material].20,33,5663,65 The operational stability of the quasi-CW laser system was demonstrated in Fig. S30 in the Supplementary Material, maintaining stable lasing intensity under ns-pumping at 2Pth for 1 h. High-quality single-mode lasers can also be achieved under fs-pumping, exhibiting a low threshold of 9.9  μJcm2 and a high Q factor of 3600 [Figs. S31(a) and S31(b) in the Supplementary Material]. These results underscore the potential of w/PEAAc films as an optical gain medium toward CW lasing or even electrically driven lasing.

    Single-mode vertical cavity surface-emitting laser. (a) Schematic of the vertical cavity surface-emitting laser. (b) Evolution of PL spectra under various pump fluences under quasi-continuous ns-pumping. Inset: far-field pattern of the lasing. (c) Integrated intensity as a function of pump fluence. Inset: lasing spectrum with a narrow FWHM of 0.14 nm, indicating a Q factor of 3850. (d) Comparison of lasing thresholds and quality factors of perovskite lasers under ns-pumping.20" target="_self" style="display: inline;">20,33" target="_self" style="display: inline;">33,56" target="_self" style="display: inline;">56–63" target="_self" style="display: inline;">–63

    Figure 5.Single-mode vertical cavity surface-emitting laser. (a) Schematic of the vertical cavity surface-emitting laser. (b) Evolution of PL spectra under various pump fluences under quasi-continuous ns-pumping. Inset: far-field pattern of the lasing. (c) Integrated intensity as a function of pump fluence. Inset: lasing spectrum with a narrow FWHM of 0.14 nm, indicating a Q factor of 3850. (d) Comparison of lasing thresholds and quality factors of perovskite lasers under ns-pumping.20,33,5663" target="_self" style="display: inline;">63

    3 Conclusion

    We have employed a volatile ammonium-driven phase reconstruction strategy to develop a high-performance gain medium with suppressed Auger recombination, achieving single-mode lasing characterized by a low threshold and a high Q factor under quasi-CW excitation. The volatile PEAAc was introduced into the all-organic perovskite system, interacting with CsPbBr3 in the precursor and desorbing during thermal annealing. This process induces a phase reconstruction that enhances the crystallinity and coverage of the polycrystalline films. More importantly, the crystallization regulated by PEAAc eliminates low-dimensional phases, thereby significantly suppressing Auger recombination. Based on this, the PEAAc-modified films exhibit a low ASE threshold of 9.6  μJcm2, along with good ASE stability under quasi-continuous nanosecond-pulsed excitation. Ultimately, by integrating the perovskite film with a pair of DBRs, we constructed a vertical-cavity surface-emitting laser with a low threshold of 17.3  μJcm2 and a high Q factor of 3850 under quasi-CW excitation. This study not only advances the performance envelope of quasi-CW lasing but also underscores the critical role of Auger recombination suppression in achieving CW lasing and even electrically driven lasing.

    Acknowledgments

    Acknowledgment. This work was financially supported by the “Pioneer” and “Leading Goose” R&D Program of Zhejiang (Grant No. 2024C01191), the National Natural Science Foundation of China (Grant Nos. U22A20133 and 52072337), the Fundamental Research Funds for the Central Universities (Grant No. 2024QZJH10, X.D.), and the Shanxi-Zheda Institute of Advanced Materials and Chemical Engineering (Grant No. 2022SZ-TD004, H.H.).

    Xinyang Wang received his BS degree from Zhejiang University in 2022. He is currently a PhD candidate at Zhejiang University. His research interests focus on materials and devices for perovskite-based lasers.

    Guochao Lu received his PhD from Zhejiang University in 2024. His research interests include optical gain mechanisms and blue VCSEL device development in perovskite films.

    Jing Li received his PhD from Zhejiang University in 2020. He is currently an associate professor at Zhejiang University of Technology. His research interests focus on understanding the fundamental properties of optoelectronic materials and advancing their device applications.

    Xingliang Dai received his PhD from Zhejiang University in 2017. He is currently a professor at Zhejiang University. His research interests focus on quantum dot emissive materials as well as their photonic and optoelectronic devices.

    Haiping He received his PhD from University of Science and Technology of China in 2004. He is currently a professor at Zhejiang University. His research interests include photophysical properties of semiconductor materials and their applications in photoluminescent/electroluminescent devices.

    Biographies of the other authors are not available.

    Tools

    Get Citation

    Copy Citation Text

    Xinyang Wang, Guochao Lu, Qiuting Cai, Jing Li, Haoran Zhang, Zaishang Long, Meiyi Zhu, Yun Gao, Qingli Cao, Hanyan Huang, Xingliang Dai, Zhizhen Ye, Haiping He, "Volatile ammonium-driven perovskite phase reconstruction for high-performance quasi-CW lasing," Adv. Photon. 7, 056006 (2025)

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Save article for my favorites
    Paper Information

    Category: Research Articles

    Received: Mar. 27, 2025

    Accepted: Jul. 24, 2025

    Published Online: Aug. 19, 2025

    The Author Email: Jing Li (lijing23@zjut.edu.cn), Xingliang Dai (shanfeng@zju.edu.cn), Haiping He (hphe@zju.edu.cn)

    DOI:10.1117/1.AP.7.5.056006

    CSTR:32187.14.1.AP.7.5.056006

    Topics