1. Introduction

In recent years, all inorganic cesium lead halide perovskite (CsPbX₃, X = Cl, Br, I) nanocrystals have received intense attention because of their prominent optoelectronic properties such as tunable absorption and photoluminescence (PL) in the whole visible spectral range, large absorption cross-section, large defect tolerance and low defect density, and high photoluminescence quantum yield [1–5], making them promising for light-emitting diodes [6–10] and lasing applications [11–16]. Since the first report on the amplified spontaneous emission (ASE) from CsPbBr₃ QD thin film under femtosecond laser and nano-second laser excitation [11], many inorganic micro- and nano-perovskite laser devices such as perovskite QD lasers [12,13], nanowire lasers [14], nanoribbon lasers [15], and single-mode laser in CsPbX₃ submicron sphere [16] have been demonstrated, revealing the great potential of lasing applications. However, due the sensitivities of CsPbX₃ nanocrystals to polar molecules and light irradiation, gain properties of these CsPbX₃ nanocrystals will degrade during operation and development of high quality laser devices based on CsPbX₃ nanocrystals still requires highly stable CsPbX₃ nanocrystals with proper encapsulation.

Glasses doped with active ions or nanocrystals have long been employed as gain media for lasers and amplifiers [17,18]. Glasses containing CsPbX₃ nanocrystals can be potential candidates for laser devices. Since the first precipitation of CsPbBr₃ nanocrystals in phosphate glasses [19], CsPbX₃ nanocrystals with tunable compositions and optical properties have been intensively investigated in glasses [20–29], and stabilities of CsPbX₃ nanocrystals against harsh environment [20] and intense irradiation [21] have been significantly improved due to the inert and dense structure of glasses. Thermal- and photo-induced degradation of CsPbX₃ nanocrystals in glasses can be easily recovered through low-temperature annealing [21,29]. These particular properties of CsPbX₃ nanocrystals embedded glasses make them promising for light-emitting diodes [20–23], X-ray scintillation [28,29], and gain media [20,30–34]. The first gain property of CsPbX₃ nanocrystals embedded glasses is illustrated by the low-temperature random up-converted lasing from CsPbBr₃ nanocrystals embedded tellurite glasses [20]. Since then, amplified spontaneous emission (ASE) from CsPbX₃ nanocrystals embedded glasses have been widely reported, and ASE in the visible spectral range has been achieved by adjusting the composition and size of the CsPbX₃ nanocrystals precipitated in the glasses (Table S1) [20,30–35]. ASE from CsPbX₃ nanocrystals embedded glasses is mainly measured using variable stripe method [31–34] and focused laser beam [20,30,35]. On the one hand, it has been reported that in disordered systems, ASE resulting from the amplification of spontaneous emission by stimulated emission harnesses optical gain provided via light scattering that is induced by intrinsic disorder in the medium before light output [36]. As a result, most of the CsPbX₃ nanocrystals doped glasses showing ASE are translucent [20] or crystallized [35]. However, there are also several works on ASE from the transparent glasses containing CsPbX₃ nanocrystals [23,24]. Even though ASE is realized, the low transparency of the translucent or crystallized glasses makes it impossible to achieve lasing. Therefore, understanding of the origin of ASE from CsPbX₃ nanocrystals doped glasses is still essential for the development of gain medium. On the other hand, PL QYs of CsPbX₃ nanocrystals embedded glasses are not closely related to their ASE. Defects present in or on the surface of CsPbX₃ nanocrystals in glasses will trap the photo-generated charge carriers and decrease the PL QY, which will deteriorate the gain performance of CsPbX₃ nanocrystals. However, most of the ASE phenomena are reported from CsPbX₃ nanocrystals embedded glasses with PL QYs much lower than 50% [20,30–35], except several reports from glasses containing CsPbBr₃ nanocrystals with PL QY higher than 50% [20,23]. In addition, most of the ASE from CsPbX₃ nanocrystals embedded glasses are recorded using femtosecond laser as pumping through two-photon absorption (TPA) process, and however, the highly intense excitation of femtosecond laser can easily destroy the CsPbX₃ nanocrystals in glasses, leading to the amorphization of these nanocrystals and loss of the photoluminescence properties [21,28,37]. For example, Hu et al. reported that CsPbBr₃ nanocrystals and CsPbBr₂I nanocrystals in glasses can be amorphized by fs laser with pulse energy of 456 nJ, pulse width of 188 fs, and repetition rate of 1 kHz [37]. However, ASE observed from CsPbX₃ nanocrystals in glasses are mostly recorded with fs laser with even higher pulse energies and smaller pulse width, where these CsPbX₃ nanocrystals can be easily damaged by the fs laser. Therefore, these issues still need to be addressed.

In this work, amplified spontaneous emission from glasses containing CsPbBr₃ nanocrystals and inclusions is investigated. By controlling the glass preparation conditions, glasses containing CsPbBr₃ nanocrystals and CsPbBr₃ related inclusions are prepared. Using focused femtosecond (fs) laser excitation, no amplified spontaneous emission can be observed from the region containing CsPbBr₃ nanocrystals, due to the lack of structural inhomogeneity and weak structural resistance to fs laser irradiation. While, strong amplified spontaneous emission is realized from the inclusions containing CsPbBr₃ phases, and these inclusions exhibit much stronger resistance to fs laser irradiation. These results show that cesium lead halide perovskite phases embedded glasses is promising for optoelectronic applications.

2. Experimental

Glass with nominal composition of 19SiO₂-35B₂O₃-6Al₂O₃-8ZnO-5CaO-2Cs₂O-4K₂O-4PbO-16KBr (in mol.%) was prepared using the conventional melt-quenching and subsequent heat-treatment methods. High-purity (99.9%) chemical powders (Sinopharm Chemical Reagent Co., Ltd) were weighted and mixed thoroughly to form homogeneous batch. These mixed powders were transferred into alumina crucibles and melted in high-temperature muffle furnace. In order to obtain proper number and size of the inclusions in the final glass obtained, the glass powders were melted at optimized condition, i.e., 1200 °C for 20 min (for further information, see Note S1 and Fig. S1 in Supplement 1). After melting, the glass melt was poured onto preheated (350 °C) brass mold and pressed with another plate for quenching and solidification of the glass melt. During quenching process, the liquid phases containing cesium, lead, and bromine elements were solidified (Supplement 1, Fig. S1). The quenched glass was annealed in muffle furnace at 350 °C for 2 h to release the thermal stress built during the quenching process, and then cooled down to room temperature with the furnace by switching off the electric power. The annealed glass was further heat-treated at 480 °C for 10 h to precipitate CsPbBr₃ nanocrystals in glass. The heat-treated glass was polished with an optical finish to further characterization.

X-ray diffraction pattern of the as-annealed heat-treated glasses were recorded with X-ray diffractometer (XRD, D8 Advance, Germany). Raman spectra of the as-prepared and heat-treated glasses were recorded with a laser confocal Raman spectrometer (LabRam HR Evolution, Horiba Jobin Yvon, France) equipped with a CCD detector and 633 nm laser (30 mW) as the excitation source. Morphology and composition of phases present in the inclusions were analyzed using filed-emission scanning electron microscope (FE-SEM, Ultra Plus, Zeiss, Germany) equipped with energy dispersive spectrometer (EDS, X-Max 50, Oxford Instruments, UK). Absorption spectra of the as-prepared and heat-treated glasses were recorded using an UV/Vis/NIR spectrophotometer (Lambda 750s, PerkinElmer, USA). Photoluminescence (PL) spectra of heat-treated glass was recorded with a fluorescence spectrometer (FL3-22, Jobin-Yvon, USA) with 460 nm light as excitation. PL QY was measured by a UV-NIR quantum yield spectrometer (C13534-11, Quantaurus-QY Plus, Hamamatsu, Japan) equipped with a charge coupled device (CCD) camera, and the uncertainty in the measured PL QY is ±1%. 1030 nm femtosecond (fs) laser (Carbide-40 W, Light Conversion, Lithuania) with a repetition rate of 1 kHz and pulse width of 190 fs was used as excitation source to record the amplified spontaneous emission (ASE) spectra. In addition, 500 nm, 600 nm, 700 nm, 800 nm, and 900 nm fs lasers from optical parametric amplifier (OPA, Orpheus, Light Conversion, Poland) pumped by the 1030 nm fs laser were also used to measure the ASE spectra. For the measurement of ASE spectra, the heat-treated glass was mounted on one computer-controlled translation stage in order to adjust the position of the interaction volume between laser beam and heat-treated glass. For ASE spectra measured using the variable stripe length (VSL) method [31–34], fs laser beams were either focused into line shape width of 60 µm by one half cylinder lens and the length of the line-shaped beam was adjusted by one razor blade mounted on another computer-controlled linear translation stage. For ASE spectra measured using focused laser beam, the fs laser beams were also focused into the heat-treated glass using one convex lens with a focal length of 10 cm and the beam diameter at the focal spot was 70 µm. Spectra of the ASE of the heat-treated glass were recorded using one fiber optic spectrometer (EQ4000, WY Optics, China). All the above structural and optical characterizations were carried out at room temperature.

3. Results and discussion

The as-prepared glass specimen is highly transparent in the visible spectral range, and after heat-treatment at 480 °C for 10 h, the glass specimen becomes light yellow with faint green emission under normal room light illumination (Supplement 1, Fig. S2a and S2b). XRD pattern (Supplement 1, Fig. S2c) show that the as-prepared glass is amorphous and no detectable crystalline phases are present in the glass specimen. While, upon thermal treatment at 480 °C for 10 h, clear and strong diffraction peaks imposed on the broad diffraction halos are observed (Supplement 1, Fig. S2c), and these diffraction peaks can be assigned to cubic structure CsPbBr₃ crystalline phases (PDF#75-0412, space group $Pm\bar{3}m$). Based on the width of the diffraction peaks, average diameter of the CsPbBr₃ crystalline phases are calculated to be 11.5 nm using the Scherrer equation. Peaks at 70 cm^-1, 158 cm^-1, and 310 cm^-1 in the Raman spectra (Supplement 1, Fig. S2d) are consistent with those observed from CsPbBr₃ crystalline phases [38–40], further illustrating that CsPbBr₃ nanocrystals are formed in the heat-treated glass specimen. TEM image shown in Fig. S3a also confirms the precipitation of CsPbBr₃ nanocrystals in the glass specimen. Absorption spectra (Supplement 1, Fig. S2e) show that the as-prepared glass specimen is highly transparent with absorption edge at ∼350 nm and the heat-treated glass specimen shows one absorption peak at 505 nm, consistent with the photographs shown in Supplement 1, Fig. S2a. Upon 360 nm light excitation, the as-prepared glass exhibits broad band PL centering at ∼457 nm, which is arising from the Pb²⁺ ions in the glass matrix or crystals bonded with bromine ions [19,41]. The heat-treated glass specimen shows one narrow band PL at 526 nm with full width at half maximum (FWHM) of 28 nm (Supplement 1, Fig. S2f) and absolute PL QY of 24.8% (Supplement 1, Fig. S3b).

Even though the as-prepared and heat-treated glass specimens look transparent (Supplement 1, Fig. S2a), there exists many nearly spherical inclusions with diameters ranging from several micrometers to several hundred micrometers (Fig. 1(a) and 1(c)). To precipitate CsPbX₃ NCs in glasses, concentrations of halide compounds are relatively high in order to compensate the evaporation loss of halide during melting and to maintain high level of saturation of halide elements in the glasses. The excessive halides, which are not dissolved into the glass melt, can form the droplets and solidify during the quenching processes. Amount of the inclusions in the as-prepared glasses increases when the melting temperature is low or melting time is short. These inclusions in the as-prepared glass specimen exhibit weak blue emission (Fig. 1(b)) and the inclusions in the heat-treated glass specimen exhibit strong green emission (Fig. 1(d)). During heat-treatment, cesium, lead, and halide elements dissolved in the glasses will go through liquid-phase separation and form NCs at super-cooled state [38], leading to the green emission in the inclusions-free regions of the heat-treated glass specimen.

 

Fig. 1. Microscope images of (a,b) as-prepared and (c,d) heat-treated glass specimens recorded under warm white light (left) and 365 nm UV light (right) illumination. Scale bars in these images are 100 µm.

Download Full Size | PDF

SEM image (Supplement 1, Fig. S4a) shows that the inclusion is neither dense and nor structurally uniform, and the inclusion is rich in Cs, Pb, Br, K, Zn, Al, and B, but deficient in Ca, Si, and O (Supplement 1, Fig. S4). EDS analysis also confirms the presence of these above elements in the inclusion (Supplement 1, Fig. S4l). In the initial inclusions, the Pb-Br bond related species generate the weak blue emission upon UV light illumination (Fig. 1(a)) [19,21,41]. Upon heat-treatment, Cs, Pb, and Br in the inclusions can also form CsPbBr₃ crystalline phase and give green emission (Fig. 1(d)). Formation of CsPbBr₃ NCs in the inclusions are further confirmed by the Raman spectra (Fig. 2). A 29.4 µm-sized inclusion (Fig. 2(a)) in the heat-treated glass specimen, which shows bright green emission under 365 nm light excitation (inset in Fig. 2(a)), is characterized using Raman spectra. Raman spectra recorded at different position on x-axis are shown in Fig. 2(b). Raman peaks at 75 cm^-1 (vibration of ${[{PbB{r_6}} ]^{4 - }}$ octahedron [38,39]), 127 cm^-1 (motion of $C{s^ + }$ cations [38,39]), 185 cm^-1 (vibration mode of $KBr$ [42]), and 310 cm^-1 (second order combination of transverse and longitudinal optical phonon of CsPbBr₃ crystal) [38–40]) are observed from the Raman spectra recorded from the central part of the inclusion, confirming the precipitation of CsPbBr₃ crystalline phases and KBr crystalline phases in the inclusion upon heat-treatment. Based on these Raman peaks, Raman mappings of the inclusion are carried out in the range of 58-100 cm^-1 (Fig. 2(c)), 116-142 cm^-1 (Fig. 2(d)), 178-194 cm^-1 (Fig. 2(e)), and 274-349 cm^-1 (Fig. 2(f)). These mapping results show that CsPbBr₃ and KBr crystalline phases coexist in the central region of the inclusion. Therefore, multiple crystalline phases are formed in the inclusions.

 

Fig. 2. (a) Optical image of one inclusion in the heat-treated glass specimen using a microscope. Inset shows the optical image of the inclusion recorded under 360 nm light excitation. Diameter of the inclusion is 29.4 µm, and the inclusion is located at 27 µm below the surface of the specimen. X and Y axis shown in the figure represent the Raman scanning direction. (b) Raman spectra of the inclusion recorded at different position along the X-axis direction. Peseudo-color plots of the Raman mappings in the range of (c) 58-100 cm^-1, (d) 116-142 cm^-1, (e) 178-194 cm^-1, and (f) 274-349 cm^-1.

Download Full Size | PDF

Since both the volume fraction of CsPbBr₃ NCs in the heat-treated glass specimen and PLQY of the heat-treated glass specimen are comparable or even higher than those employed to realize the ASE [20,30–35], PL spectra of the heat-treated glass specimen is also recorded upon excitation by one 1030 nm fs laser stripe via TPA process. Firstly, the PL spectra are recorded from the inclusion-free region (Fig. 3(a) and 3(b)). The peak PL intensity and FWHM of the PL spectra do not exhibit typical feature of ASE (Fig. 3(a) and 3(b)). The peak PL intensity initial reaches the first maximum when the energy density of the laser stripe reaches 1.3 mJ/cm², decreases with further increase in energy density of the laser stripe to 3.2 mJ/cm², and gradually increases to a plateau with continuous increase in energy density of the laser stripe to 8.0 mJ/cm² (Fig. 1(b)). At the same time, FWHM of the PL spectra gradually decrease from ∼18 nm to ∼16 nm, much larger than those (<6 nm) observed from the ASE of CsPbBr₃ NCs [20,30–34]. Both changes in the PL peak intensity and FWHM of the PL spectra demonstrate that no ASE is generated from CsPbBr₃ NCs in the inclusion-free regions. The unusual dependence on PL intensity on the energy density of fs laser stripe is probably associated with the photo-stability of the CsPbBr₃ NCs in glasses. It has been reported that CsPbBr₃ NCs in glasses can also be easily damaged or amorphized by intense light beams [21,37], leading to the quenching of photoluminescence. To confirm this phenomenon, the heat-treated glass specimen is scanned by focused fs laser (Fig. 3(c)). It can be visually observed that when the energy density of focused fs laser increases to 4.1 mJ/cm² or above, the irradiated region starts to lose the PL, confirming the photo-induced damage to the CsPbBr₃ NCs in glasses (Fig. 3(c)). PL spectra recorded using the fs laser stripe or focused fs laser from many other inclusion-free regions have similar features, indicating that ASE cannot be realized in these inclusion-free regions. Similar PL spectra are also measured using the high-quality CsPbX₃ NCs embedded glasses (which do not have any bubbles or inclusions) as previously reported [19,21,22]. As a result, it is not surprising that no ASE is observed in the inclusion-free region, which is mainly induced by the low photodamage threshold of CsPbBr₃ nanocrystals due to the ionic structure and low melting temperature. However, the structural disorder or compositional inhomogeneity induced light scattering [36] can facilitate the realization of ASE in translucent or crystallized glasses containing CsPbX₃ NCs [30–33,35].

 

Fig. 3. (a) PL spectra, (b) peak intensity and full width at half maximum of PL bands recorded from heat-treated glass specimen excited by 1030 nm fs laser stripe (stripe shape: 5000 × 60 µm²) with different energy densities. (c) Focused 1030 nm fs laser scanning induced optical damages to the heat-treated glass specimen. The fs laser with repetition rate of 1 kHz is focused onto the surface the heat-treated specimen by one convex lens and the focal point is 70 µm, and the laser scanning speed is 1 mm/s. The image is recorded using a microscope upon 365 nm light illumination.

Download Full Size | PDF

On the other hand, ASE is easily observed from the inclusions in heat-treated glass specimens when they are excited by focused fs laser beam. To illustrate the ASE of the inclusions in the heat-treated glass specimen, the 29.4 µm sized inclusion (located at 27 µm below the glass surface) is excited by focused fs laser (Fig. 4). As the energy density of the laser beam increases from 0.41 to 14.13 mJ/cm², intensity of the PL band monotonically increases with simultaneous reduction in bandwidth (Fig. 4(a)). With the increase in energy density, peak intensity of PL band exhibit one inflection point at 3.9 mJ/cm², and FWHM of the PL bands decreases from ∼12 nm to ∼5 nm (Fig. 4(b)). These features are typical for ASE, indicating the ASE is realized from the inclusion in the heat-treated glasses. ASE is also mapped across the entire inclusion with focused fs laser with energy density of 10.39 mJ/cm² (Fig. 4(c) and 4(d)). Both mapping results of PL peak intensity (Fig. 4(c)) and FWHM (Fig. 4(d)) demonstrate that ASE can be realized when the excitation light is focused into the inclusion. ASE is also observed from inclusions with different sizes (Supplement 1, Fig. S5). In addition, CsPbBr₃ crystalline phase in these inclusions are much more stable than CsPbBr₃ NCs in the glass matrix. CsPbBr₃ NCs in the inclusion-free region of the heat-treated glass specimens can be damaged and become nonluminescent upon scanning of focused fs laser beam with energy density of 20.8 mJ/cm² (the black line in Supplement 1, Fig. S6). While after multiple scanning, the inclusion in the heat-treated glass specimens still remains luminescent (Supplement 1, Fig. S6). The photo-stability of the inclusions in the heat-treated glass specimen make it possible to obtain stable ASE. This feature is further demonstrated by successive measurement of ASE from the same inclusion with diameter of ∼31 µm. For the first measurement, ASE is observed when the energy density of the fs laser increases to 5.31 mJ/cm² or above (Supplement 1, Fig. S7a and S7b). For the second measurement, ASE is also observed when the energy density of the fs laser decreases from 9.33 to 1.33 mJ/cm², and the inflection point is observed at 5.18 mJ/cm² (Supplement 1, Fig. S7c and S7d), slightly smaller than that observed during the first measurement. Compared with the first measurement, spontaneous emission intensity of the inclusion is lower, which is induced by the fs laser irradiation induced damage. ASE measured in the third and fourth measurements (Supplement 1, Fig. S7e-S7h) are almost identical to that observed in the second measurement. Results shown in Supplement 1, Fig. S6 and Fig. S7 demonstrate that inclusion in glass is stable and suitable for application in gain medium.

 

Fig. 4. PL spectra of one 29.4 µm sized inclusion located at 27 µm below the surface of the heat-treated specimen excited by focused 1030 nm fs laser. (a) PL spectra recorded with different energy densities, (b) changes in PL peak intensity and FWHM with energy densities. (c) PL peak intensity and (d) FWHM of PL spectra recorded with energy density of 10.39 mJ/cm² across the entire inclusion.

Download Full Size | PDF

ASE of these inclusions in glasses also exhibits strong dependence on the excitation wavelength (Fig. 5). Taken one inclusion with diameter of 140 µm as example (Fig. 5(a)), ASE is observed when the wavelength of the fs laser is tuned from 500 nm to 1030 nm (Supplement 1, Fig. S8). The threshold energy density of the focused fs laser for ASE increases from 1.22 mJ/cm² to 15.9 mJ/cm² when the wavelength of the fs laser is tuned from 500 nm to 1030 nm (Fig. 5(b)). Increase in threshold values for ASE with the increase in wavelength of fs laser is closely related to the reduced two-photon absorption cross section [43].

 

Fig. 5. (a) Optical image of one inclusion with diameter of 140 µm (b) threshold of energy density for ASE recorded with fs lasers of different wavelengths.

Download Full Size | PDF

To further evaluate the gain properties of these inclusions, PL spectra of one inclusion with diameter of ∼50 µm are investigated using the VSL method, which is schematically illustrated in Supplement 1, Fig. S9. For this measurement, 1030 nm fs laser with repetition rate of 1 kHz is shaped into stripe by one cylinder lens, and energy density of the laser stripe at the focal is fixed at 6 mJ/cm². The length of the laser stripe is ∼5 mm, and the stripe length with interaction with the glasses is gradually tuned to 2 mm by blocking the extra with a razor blade. When the stripe length is shorter than 2.4 mm, only spontaneous emission (SE) is observed, which is characterized by the low PL intensity and large FWHM (Fig. 6(a) and 6(b)). When the stripe length is between 2.9 mm and 5.0 mm, both SE and ASE are observed (Fig. 6(a) and 6(b)). With the reduction in stripe length from 5.0 mm to 2.9 mm, PL peak intensity gradually decreases and FWHM is maintained at ∼4.8 nm (Fig. 6(b)). Decrease in PL peak intensity is mainly caused by the reduction in SE from the CsPbBr₃ NCs in glasses due to the reduced excitation volume by the shortening of the laser stripe (Fig. 6(a)). When the length of the laser stripe is further reduced from 2.9 to 2.5 mm, the PL peak intensity slightly increases and the FWHM decreases to ∼4.2 nm (Fig. 6(b)). The slight increase in PL peak intensity is assigned to the reduced reabsorption of ASE by the inclusion, which is consistent with the slight blue shift of the ASE peak (Fig. 6(a)). With further reduction in laser stripe length from 2.5 mm, PL peak intensity exhibits rapid drop with abrupt increase in FWHM, and ASE disappears when laser strip length is shorter than 2.4 mm (Fig. 6(a) and 6(b)). To evaluate the gain performance of the inclusion, the PL peak intensity recorded with a laser stripe length between 2.4 mm and 2.5 mm is fitted using the equation ${I_0}(L )= ({{I_S}A/g} )[{\exp ({gL} )- 1} ]$ (where $L$ is laser stripe length, ${I_0}(L )$ is the PL peak intensity, A represents the excited area, ${I_S}$ represents the spontaneous emission rate per unit volume, $g = g^{\prime} - \alpha $ (cm^-1) represents the net gain under the excitation, $g^{\prime}$ is the gain due to the stimulated emission, and $\alpha $ is the optical loss) [44,45]. Fitting with the equation (Fig. 6(c)) yields a net gain of 409 ± 47 cm^-1 at room temperature. This net gain is comparable to those recorded from CsPbBr₃ NCs [46,47]. The high photo-stabilities and large gain values of the inclusions in the glasses indicates they are potential gain materials.

 

Fig. 6. (a) PL spectra, (b) PL peak intensity and FWHM of one inclusion with diameter of 50 µm in the glass excited by 1030 nm laser stripe with different length. Energy density of the laser stripe at the focal is fixed at 6 mJ/cm², and the length of the laser stripe is shortened from 5 mm to 2 mm by adjusting the position of the razor blade. Relative position of the inclusion in the laser stripe is indicated by the green shadow in (b). (c) Fitting of the PL peak intensity with stripe length in the range of 2.1 mm to 2.7 mm.

Download Full Size | PDF

4. Conclusion

In summary, ASE of CsPbBr₃ nanocrystals embedded in glasses is investigated. It is found that CsPbBr₃ nanocrystals dispersed in glass matrix can be damaged by the intense fs laser pulses and cannot support the ASE. ASE is found to originate from CsPbBr₃ nanocrystals in the inclusions formed in the glass. ASE is observed from CsPbBr₃ nanocrystals containing inclusions with sizes in the range of several tens of micrometers to several hundred micrometers. These CsPbBr₃ nanocrystals in the inclusions are more resistant to fs laser irradiation and maintain stable ASE, and ASE can still be observed when the pulse energy density is as high as several tens of mJ/cm². Threshold values of energy densities for the ASE increases with the increase in fs laser wavelength. High optical gain is achieved from these inclusions containing CsPbBr₃ nanocrystals, and net optical gain as high as 409 ± 47 cm^-1 is realized from one ∼50 µm-sized inclusion. These results demonstrate that glasses containing CsPbBr₃ nanocrystals embedded inclusions can be engineered for potential gain applications.