Cesium lead halide (Cl, Br, I, or halide mixture) is considered to be a promising photoelectronic material, including employment in a light-emitting diode (LED), solar cell, laser, biomedical detection, and so on^[1–5]. All of these superiorities contribute to their outstanding optical properties, such as facile controllable photoluminescence (PL) emission spectra, narrow full width at half-maximum (FWHM) of emission spectra and high PL quantum yield (PLQY)^[1]. However, it is a major problem to maintain the optical properties of perovskite quantum dots (QDs) in air. Researchers have proposed a series of strategies to eliminate the defect of perovskite QDs. Actually, the QDs could still maintain outstanding optical properties after being treated by ion doping^[6–14], surface passivation^[15], and hybrid construction^[16–33]. Therefore, balancing the stability and optical properties of perovskite QDs has become the focus of research.

Encapsulating QDs into other media has been considered an efficacious method to avoid their fluorescence degradation. Many facile methods were proposed to embed QDs into other materials to form cladding structures^[18–29] and core-shell structures^[16,17]. Actually, silica is an excellent choice to protect core materials because of its high stability and pellucidity. Besides, compounds of QDs/silica substrates can be simply fabricated into fluorescence powder and luminescence film. Whereafter, all kinds of QDs/silica structures were fabricated, including mesoporous silica^[20,22], QDs/SiO2 spheres^[20,27,29], and QDs/SiO2 Janus nanocrystals^[33]. Silane reagents are commonly used to obtain silica, such as (3-aminopropyl)trimethoxysilane (APTMS)^[6], (3-aminopropyl)triethoxysilane (APTES)^[23,26,29], perhydropolysilazane (PHPS)^[25], tetraethyl orthosilicate (TEOS)^[34], and tetramethyl orthosilcate (TMOS)^[28,35]. Especially, APTES is considered a typical silane reagent to form silica because of the extensive usage and budget price.

In this work, we synthesize CsPbBr3 perovskite QDs at room temperature firstly^[29,36]. Then, we use the anion exchange method to replace Br− with Cl− in a short time. At the same time of ion exchange reaction, APTES is gradually hydrolyzed into SiO2. Finally, we can obtain CsPb(Cl/Br)3/SiO2 nanospheres with different emission wavelengths by coordinating the ion exchange time and chloride concentration^[37–39]. The final product also has humidity stability and thermal stability. This efficient cyan CsPb(Cl/Br)3 QD does make up for the lack of efficient cyan commercial phosphors.

As for the experimental process of synthesizing CsPb(Cl/Br)3@SiO2 spheres, first, 0.326 g (1 mmol) Cs2CO3 was dissolved in 5 mL n-octanoic acid (OTAc), keeping the solution stirring for 10 min to obtain the Cs precursor. Then, 0.367 g (1 mmol) PbBr2 and 1.094 g (2 mmol) tetraoctylammonium bromide (TOAB) were added into 3 mL toluene, while the mixture was stirred for 15 min. By injecting 0.7 mL APTES into the PbBr2-toluene and stirring for 20 s, 1.0 mL Cs precursor is rapidly added. After that, 180 μL LiCl-H2O (2.01 g LiCl in 3 mL deionized water) was injected into the mixture. Whereafter, the precipitate was separated by centrifugation after 10 min of stirring. The deposit was dispersed in n-hexane, and the crude solution was recentrifuged. Finally, the as-prepared compound solid needed to be dried for 48 h at 40°C to obtain QDs/SiO2 powder.

The fabrication schematic of the CsPb(Cl/Br)3/SiO2 mixture solution is demonstrated in Fig. 1. In order to preserve and employ the luminous materials more conveniently, we extracted the QDs/SiO2 compound from the sol-gel solution via centrifugation. Afterwards, we ground the drought-treated solid compound to luminescence powder. Actually, there are small facade changes when 180 μL of LiCl-H2O solution was injected into CsPbBr3/APTES mixture firstly. But, previously, the color of the mixture changes from yellow–green to cyan gradually. The process of color transformation was completed in about 10 min. This phenomenon indicates that the ion exchange reaction is fast and controllable. More importantly, the final samples emit intense cyan luminescence under ultraviolet (UV) light. Therefore, it is necessary to explore the reaction process and optical properties of the QDs/SiO2 composite.

We performed a series of transmission electron microscopy (TEM) images of the final QDs/SiO2 composite with different magnification. As shown in Fig. 2(a), many QDs/SiO2 spheres were evenly distributed. Dispersed QDs/SiO2 spheres also have similar dimensions. We calculated the size distribution of the CsPb(Cl/Br)3/SiO2 nanospheres [inset of Fig. 2(a)]. The average diameter of the nanospheres is 4.95 nm, while the maximum and minimum diameters are 9.38 nm and 2.15 nm, respectively. Besides, nanospheres in the 3.6 nm to 6.8 nm range accounted for 87% of the total statistics. We hypothesized that the silane reagent is responsible for the binding of particles. In this experiment, APTES is indeed involved in the combination of Pb2+, Br-, and Cs+. In essence, APTES controls the growth of QDs. There are also many QDs attached to amorphous silica substrates [Fig. 2(b)] in our product. From Fig. 2(b), it can be observed that undispersed composite still maintains the shape of the spheres. To verify whether the mixed phases were formed by ion exchange, we measured the spacing of the lattice stripes and matched the results with the database. Three different values of 0.290 nm, 0.252 nm, and 0.278 nm correspond to the crystal faces of (2 0 0) (CsPbBr3, JPCDS No. 54-0752), (2 1 0) (CsPbCl3, JPCDS No. 75-0411), and (2 0 0) (CsPbCl3, JPCDS No. 75-0411) in Fig. 2(c), respectively. Besides, selected area electron diffraction (SAED) of the nanoparticle indicates that final QDs product has high degrees of crystallinity [inset of Fig. 2(c)]. The above TEM images demonstrated the successful synthesis of CsPb(Cl/Br)3@SiO2 spheres.

We fabricated a series of samples by dropping different amounts of LiCl solution, while other experimental conditions remained exactly the same. After the calculation, 60 μL, 100 μL, 140 μL, and 180 μL of LiCl-H2O correspond to 0.95 mmol, 1.58 mmol, 2.21 mmol, and 2.84 mmol of Cl-, respectively. The specific ion content provides a basis for us to judge the degree of anion exchange. As depicted in Fig. 3(a), with the increase of the chloride dose, the emission wavelength decreased from 519 nm to 487 nm. Corresponding to Fig. 3(c), the color changed from green to cyan while the samples were exposed under UV irradiation (λ=365nm) [Fig. 3(c)]. Combining the PL spectra and the standard color spectrum, it can be speculated that the different concentrations of the Cl anion could prepare different colors of QDs product. In other words, we could adjust the amount of LiCl-H2O to obtain tunable PL emission spectra. But, the PL intensity of the five samples has a tendency to go down first and then rise with the increase of Cl anion concentration [Fig. 3(b)]. It can be inferred that the Cl anion could gradually displace the Br anion in situ to form CsPb(Cl/Br)3 within the initial period. As the amount of LiCl-H2O rises from 140 μL to 180 μL, Cl ions dominate the perovskite halogen position that replaces Br ions. Besides, the weight of the sample ranges from 0.9 g to 1.0 g in our experiment. In the field of synthesizing CsPbX3 QDs powder, a maximum yield of 1.0 g is indeed a high yield.

To verify our speculation, X-ray diffraction (XRD) equipment was utilized to characterize the composition of the QDs/SiO2 composite [Fig. 4(a)]. The real peaks exhibit an obvious shift to higher angles compared with the CsPbBr3 (JPCDS No. 54-0752) standard card, while the actual peaks display lower angles compared with CsPbCl3 (JPCDS No. 75-0411). Combined with the overall peak variation trend, it can be learned that the composition of the material changes from the CsPbBr3 phase [crystal structure of cubic CsPbBr3 is shown in Fig. 4(b)] to the CsPbCl3 phase. Combining PL spectra and XRD patterns, we can infer that CsPb(Cl/Br)3 is generated. Besides, we found that the remaining peaks (12.6°, 20.1°, and rest of the small peaks) correspond to rhombohedral Cs4PbBr6 (JPCDS No. 73-2478). In our previous work, we proved that the generation of rhombohedral Cs4PbBr6 has no negative effect on optical properties of cubic CsPbBr3^[27]. Here, we can only make a preliminary guess at the cause of the increase in deionized water. The increase of water molecules breaks the ratio of Cs, Pb, and Br, resulting in the production of Cs4PbBr6.

Besides, we acquired the PL excitation (PLE) spectra [Fig. 5(a)] of the QDs/SiO2. It can be found that the QDs behave in the broad excitation range from 350 nm to 460 nm. Besides, the similar trend of the curves indicates that the five samples have a similar essence. The PLE spectrum indicates that QDs/SiO2 powder has uniform absorption ability from UV light to blue light. The immediate merit is that CsPb(Cl/Br)3/SiO2 could be employed in UV chips and blue chips indiscriminatingly. We also compared the Fourier transform infrared (FTIR) spectra of the CsPb(Cl/Br)3@SiO2 and SiO2 [Fig. 5(b)]. The SiO2 was prepared by natural hydrolysis of APTES at room temperature. The hydrolysis process occurs in a culture dish for more than 24 h without stirring. As shown in Fig. 5(b), the similarity of the two patterns indicates that the spontaneous hydrolysis of APTES has the same effect as that of accelerated hydrolysis. Besides, only several sharp peaks show the difference between two substances, including the ligands of some residual organics. Therefore, the message we can obtain from the figure is that the QDs we fabricated are indeed embedded in the silica substances.

The CIE color coordinates of CsPb(Cl/Br)3 were also calculated. As shown in Fig. 6, the coordinates of the five samples are (0.106, 0.743), (0.063, 0.653), (0.053, 0.542), (0.060, 0.381), and (0.075, 0.266), respectively. The tendency of the QDs@SiO2 changes from green to cyan indicated that the method we proposed does have the possibility to achieve blue emission pure CsPbCl3@SiO2 at room temperature. Actually, the CsPbBr3@SiO2 we synthesized before only has a PLQY of ∼70%^[27]. On that basis, we achieved a PLQY of 84% after adding 60 μL LiCl-H2O solution. In fact, Cl can only partially replace Br on cubic CsPbBr3 to form CsPb(Cl/Br)3. There are still some free Cl- and Li+ in the sample. Underutilized Cl and Li produce a doping effect on CsPbBr3. A series of tests were performed to verify the stability of QDs. The PL intensity of CsPb(Cl/Br)3@SiO2 powder can almost remain 100% after 30 min when treated in moist air (∼85% humidity). Further, composites were rapidly heated to 105°C (∼30 s) and then naturally cooled to room temperature (25°C); the PL intensity of CsPb(Cl/Br)3@SiO2 powder had little variation (∼100%). The heating–cooling cycle test indicates that the luminescence characteristics of CsPb(Cl/Br)3@SiO2 are reversible. Finally, the cyan emission powder has a PLQY of 84% at the maximum, so this stable cyan CsPb(Cl/Br)3/SiO2 powder with high PLQY indeed has great potential to be employed in practice.

In summary, we successfully synthesized silica-wrapped CsPb(Cl/Br)3 perovskite QDs at room temperature in open air. The QDs/SiO2 composites were extracted from the sol-gel solution by using a precipitation-encapsulation method assisted with APTES. Cesium lead halide QDs with different wavelengths can be prepared by adjusting the concentration of the LiCl-H2O. The QDs@SiO2 composites with tunable wavelengths behave with improved stability and a high PLQY of 84%. Besides, humidity stability and thermal stability show that the CsPb(Cl/Br)3@SiO2 composite possesses great application value. The fabrication of cyan fluorescent powder indeed behaves with immense potential in a full spectrum display. The high yield of the powder is also performed with an industrial application prospect for large-scale preparation.