A Blue-Light-Emitting 3 nm-Sized CsPbBr 3 Perovskite Quantum Dot with ZnBr 2 Synthesized by Room-Temperature Supersaturated Recrystallization

: Recently, tuning the green emission of CsPbBr 3 quantum dots (QDs) to blue through quantum size and conﬁnement effects has received considerable attention due to its remarkable photophysical properties. However, the synthesis of such a blue-emitting QD has been challenging. Herein, supersaturated recrystallization was successfully implemented at room temperature to synthesize a broadband blue-emitting ZnBr 2 -doped CsPbBr 3 QD with an average size of ~3 nm covering the blue spectrum. The structural and optical properties of CsPbBr 3 QDs demonstrated that QD particle size may decrease by accommodating ZnBr 2 dopants into the perovskite precursor solution. Energy-dispersive spectroscopy conﬁrmed the presence of zinc ions with the QDs. This work provides a new strategy for synthesizing strongly quantum-conﬁned QD materials for photonic devices such as light-emitting diodes and lighting.

Normally, a mixed halide perovskite such as CsPbCl x Br 3−x can cover the entire blue spectrum [15,16]. However, this approach is constrained by lattice mismatch, phase segregation, and chlorine vacancies [17][18][19]. Specifically, the Cl vacancies are responsible for deep defects and the low PLQYs [13,17,19]. On the other hand, CsPbBr 3 has a stable structure with fewer crystal defects due to its well-fitted atomic radii [19]. However, the instability of the perovskites to photon, thermal, moisture, and operational challenges is still a bottleneck for the commercialization of perovskite QD-based devices. The instability problem is more pronounced for Cl-based QDs due to defects as well as the intrinsic distortion of the [PbCl 6 ] 4− octahedron [20], opening the channel for the degradation of materials. Moreover, mixed CsPbCl x Br 3−x perovskites are prone to both phase segregation and spectral instability [21]. To overcome these drawbacks, in situ and/or post-treatment Cl defect passivation has been investigated using versatile ligands. However, excess ligands may impede the charge-transport properties of the perovskite QDs [22]. As a result, many researchers have examined pure bromine CsPbBr 3 QDs as a source for blue emission by substituting Pb 2+ with a similar-sized metal cation considering Goldschmidt's tolerance factor, t = (r A + r X )/ √ 2 · (r B + r X ) , with 0.813 ≤ t ≤ 1.107 for structural stability, where r A , r B , and r X are the radius of A cation, B cation, and X anion in the ABX 3 perovskite structure, respectively [23]. For example, the blue shift of optical spectra was demonstrated by partially replacing Pb 2+ with a divalent cation (Cd 2+ , Zn 2+ , Sn 2+ , Cu 2+ ) and a trivalent cation (Al 3+ , Sb 3+ ) [11,[24][25][26]. Here, the blue shift is attributed to the lattice contraction as these ions possess a smaller ionic radius than Pb 2+ . As a result, the Pb-Br bond becomes shorter and increases interaction between Pb and Br orbitals, which is responsible for the blue shift [11,[24][25][26][27][28][29]. Besides emission, the dopant metallic cation can enhance the stability of the CsPbBr 3 not only by increasing the defect formation energy but also passivating the defect state of the QDs [26][27][28][29][30].
The other strategy is controlling the quantum size of CsPbBr 3 NCs in a quantumconfinement regime [28,29]. The two most common synthesis methods, i.e., hot injection (HI) and supersaturation recrystallization (SR), usually afford CsPbBr 3 QDs with a size of greater than 7 nm, resulting in green light emission. Hence, different techniques have been employed to control the CsPbBr 3 QDs in strong confinement regimes like ligand composition engineering (oleic acid/oleylamine ratio change) [31], hydrogen bromide (HBr) acid etching-driven ligand exchange [32], synthesis under thermodynamic equilibrium environment [9,33], and the two-step SR technique [34]. Importantly, to date, most of the blue-emitting CsPbBr 3 QD synthesis has been based on the typical HI method, which requires a high-temperature injection as well as an inert gas environment, limiting its practical application due to energy and material cost.
In this work, we synthesized a blue light-emitting ZnBr 2 -doped CsPbBr 3 QD using the SR method at room temperature (without any inert gas but under ambient conditions). This approach is initiated by the principle of the size-dependent stoichiometry of Br − in CsPbBr 3 QDs (higher Br − contents in the smaller QDs) and the equilibrium between the QDs and colloidal dispersion medium [33,35]. The~3 nm-sized QDs were obtained by adding a controlled amount of ZnBr 2 into the perovskite precursor solution, resulting in a broadband blue emission.

Synthesis of ZnBr 2 -CsPbBr 3 QDs
The ZnBr 2 was prepared by the SR method at room temperature [36]. Under magnetic stirring, 0.4 mmol CsBr, 0.4 mmol PbBr 2 , 1 mL OA, and 0.5 mL OAm were dissolved in 10 mL of DMF for 2 h as a source of Cs, Pb, and Br. Under the same conditions, 2.5 mmol ZnBr 2 was dissolved in 5 mL of DMF for 2 h. Then, 0 and 200 µL of a ZnBr 2 solution and 1 mL of the perovskite precursor (PbBr 2 /CsBr) solution were simultaneously injected into Photonics 2023, 10, 802 3 of 13 10 mL of toluene under vigorous magnetic stirring. The synthesized QDs were centrifuged at 3500 rpm for 5 min. Then, the QD dispersion mixed with 5 mL ethyl acetate was centrifuged at 9000 rpm for 10 min. Finally, the precipitated QDs were used for further characterization after re-dispersing in toluene acting as an antisolvent.

Computational Methods
The electronic band structures of the 3 × 3 × 3 CsPbBr 3 supercells without or with zinc doping were calculated using Cambridge Serial Total Energy Package (CASTEP, Materials Studio 2017, Vélizy-Villacoublay, France). General gradient approximation (GGA) with perdew-burke-ernzerhof (PBE) exchange-correlation functional was used to calculate both geometry optimization and the electronic properties of the materials [37]. The Mohkhorst pack grid of 3 × 3 × 3 size was constructed for k-points in the Brillouin zone. The energy of 1.0 × 10 −5 eV/atom, force of 0.02 eV/Å, maximum displacement of 4 × 10 −4 Å, and maximum stress of 0.04 GPa were used for geometry optimization. Figure 1 shows the schematic explanation of the supersaturated recrystallization process for the synthesis of CsPbBr 3 QDs. As shown in Figure 1a, when a green-lightemitting CsPbBr 3 QD was synthesized, the perovskite colloidal dispersion was transparent yellow. On the other hand, when a blue-light-emitting CsPbBr 3 QD was prepared, that was partially cloudy without any yellowish color, implying that the QD size might be very small with a strong quantum size and confinement effect in nanoscale. The desired amount of ZnBr 2 as described in the experimental section was employed to reduce the size of CsPbBr 3 QDs in nanoscale. This approach is consistent with the literature reports [33,35], in which ZnBr 2 was a source of excess Br − ions to adjust the size of CsPbBr 3 QDs. Smaller QDs have a larger surface area requiring sufficient amounts of surface ligands and/or Lewis base Br − ions.

Results and Discussion
The CsPbBr 3 QDs without/with ZnBr 2 dopant were synthesized by SR under ambient conditions [38]. Figure 2 shows the XRD patterns of the CsPbBr 3 drop-cast films [38][39][40][41], corresponding to the cubic phase of CsPbBr 3 [40]. Please note that CsPbBr 3 shows polymorphism depending on temperature, like orthorhombic (≤88 • C), tetragonal (88 ≤ T ≤ 130 • C), and cubic (≥130 • C) [38][39][40][41]. However, in the case of nanoscale crystals, the cubic phase was frequently observed in comparison to the orthorhombic, indicating the metastable state of QDs, i.e., kinetically metastable or stable but thermodynamically unstable. The XRD patterns for ZnBr 2 -doped CsPbBr 3 drop-cast film display more orientational order compared to the pristine CsPbBr 3 film without ZnBr 2 doping. Importantly, the same XRD patterns indicate that the zinc-ion doping into the CsPbBr 3 NCs does not affect the crystal structure of this perovskite crystal. This observation is consistent with the literature reports confirming that the zinc ion maintains a crystal phase of CsPbBr 3 [42,43]. However, the presence of a lot of small XRD peaks indicates that the orientational order is very small in the drop-cast CsPbBr 3 thin films composed of polydisperse nanoparticles. In this work, any post-annealing process, co-solvent addition, and antisolvent dispensing (so-called solvent engineering) were not employed, which should affect the crystal orientation during film formation via the intermediate phase of the perovskite precursors. Importantly, nanocrystals have a high surface energy and weakly bound surface ligands in the case of oleic acid, which may allow a further aggregation of QD particles in order to decrease free energies during thin-film drying processing, indicating a meta-stability of QDs. However, further detail could be addressed in our future work regarding the phase transformation and stability of nanoscale QDs. The CsPbBr3 QDs without/with ZnBr2 dopant were synthesized by SR under ambient conditions [38]. Figure 2 shows the XRD patterns of the CsPbBr3 drop-cast films [38][39][40][41], corresponding to the cubic phase of CsPbBr3 [40]. Please note that CsPbBr3 shows polymorphism depending on temperature, like orthorhombic (≤88 °C), tetragonal (88 ≤ T ≤130 °C), and cubic (≥130 °C) [38][39][40][41]. However, in the case of nanoscale crystals, the cubic phase was frequently observed in comparison to the orthorhombic, indicating the metastable state of QDs, i.e., kinetically metastable or stable but thermodynamically unstable. The XRD patterns for ZnBr2-doped CsPbBr3 drop-cast film display more orientational order compared to the pristine CsPbBr3 film without ZnBr2 doping. Importantly, the same XRD patterns indicate that the zinc-ion doping into the CsPbBr3 NCs does not affect the crystal structure of this perovskite crystal. This observation is consistent with the literature reports confirming that the zinc ion maintains a crystal phase of CsPbBr3 [42,43]. However, the presence of a lot of small XRD peaks indicates that the orientational order is very small in the drop-cast CsPbBr3 thin films composed of polydisperse nanoparticles. In this work, any post-annealing process, co-solvent addition, and antisolvent dispensing (so-called solvent engineering) were not employed, which should affect the crystal orientation during film formation via the intermediate phase of the perovskite precursors. Importantly, nanocrystals have a high surface energy and weakly bound surface ligands in the case of oleic acid, which may allow a further aggregation of QD particles in order to decrease free energies during thin-film drying processing, indicating a meta-stability of QDs. However, further detail could be addressed in our future work regarding the phase transformation and stability of nanoscale QDs.  Figure 3 shows the EDS spectra for CsPbBr3 QDs. As shown in Figure 3, when ZnBr2 was incorporated into the perovskite precursor solution, the synthesized CsPbBr3 QDs exhibited the presence of the zinc element (see Figure 3b). Importantly, the zinc ions can stay with CsPbBr3 QDs in two possibilities, i.e., substitutional (e.g., substitution of Pb 2+ ions), and interstitial substitution in the bulk and/or surface of CsPbBr3 QDs. The exact location is out of scope in the current study. However, based on literature reports [23], the zinc ions are known to stay in the CsPbBr3 crystal structure by partially substituting the Pb 2+ ions in the bromide plumbate [29].  Figure 3 shows the EDS spectra for CsPbBr 3 QDs. As shown in Figure 3, when ZnBr 2 was incorporated into the perovskite precursor solution, the synthesized CsPbBr 3 QDs exhibited the presence of the zinc element (see Figure 3b). Importantly, the zinc ions can stay with CsPbBr 3 QDs in two possibilities, i.e., substitutional (e.g., substitution of Pb 2+ ions), and interstitial substitution in the bulk and/or surface of CsPbBr 3 QDs. The exact location is out of scope in the current study. However, based on literature reports [23], the zinc ions are known to stay in the CsPbBr 3 crystal structure by partially substituting the Pb 2+ ions in the bromide plumbate [29].
at room temperature. Figure 3 shows the EDS spectra for CsPbBr3 QDs. As shown in Figure 3, when ZnBr2 was incorporated into the perovskite precursor solution, the synthesized CsPbBr3 QDs exhibited the presence of the zinc element (see Figure 3b). Importantly, the zinc ions can stay with CsPbBr3 QDs in two possibilities, i.e., substitutional (e.g., substitution of Pb 2+ ions), and interstitial substitution in the bulk and/or surface of CsPbBr3 QDs. The exact location is out of scope in the current study. However, based on literature reports [23], the zinc ions are known to stay in the CsPbBr3 crystal structure by partially substituting the Pb 2+ ions in the bromide plumbate [29].    Figure 4a revealed that the shape of the pristine CsPbBr 3 QDs is cuboidal [44,45] with an average edge length of (~22 nm), whereas that of the ZnBr 2 -doped CsPbBr 3 QDs is very small,~3 nm (Figure 4d). This TEM image demonstrates that adding the appropriate amount of ZnBr 2 dopants into the perovskite precursor solution may allow control of the size of CsPbBr 3 in the strong quantum-confinement regime, affording the blue emission in the below. Figure 5 shows the size distribution of CsPbBr 3 QDs based on the aforementioned HR-TEM images. The pristine CsPbBr 3 QDs exhibit an average QD size of~22 nm. However, the range is somewhat broad, from~7 nm to~50 nm, indicating the polydispersity of nanoparticles in this SR method at room temperature, although the chosen conditions (OA and OAm ratio) should affect this QD size and distribution. Importantly, when ZnBr 2 was employed as a dopant for CsPbBr 3 QDs, the QD size is in the range of~1.5 to~5.5 nm, with the average size of~3 nm, guaranteeing the quantum size and confinement effect, a blue-light emission in this study. Figure 4 shows the HR-TEM mages of CsPbBr3 QDs (Figure 4a,b) without ZnBr2 and (Figure 4c,d) with ZnBr2 doping, respectively. Figure 4a revealed that the shape of the pristine CsPbBr3 QDs is cuboidal [44,45] with an average edge length of (~22 nm), whereas that of the ZnBr2-doped CsPbBr3 QDs is very small, ~3 nm (Figure 4d). This TEM image demonstrates that adding the appropriate amount of ZnBr2 dopants into the perovskite precursor solution may allow control of the size of CsPbBr3 in the strong quantum-confinement regime, affording the blue emission in the below.  The optical properties of CsPbBr 3 without and with ZnBr 2 doping are summarized in Figure 6. As indicated above, the size of ZnBr 2 -doped CsPbBr 3 QDs was about 3 nm, which is much less than the Bohr exciton radius of~7 nm [14]. The very small size of the quantum dot was responsible for the significant blue shift of absorption and emission peaks (Figure 6c) due to the quantum-confinement effect. Moreover, the blue shift might be assisted by the lattice internal stress, owing to the zinc-ion doping in the perovskite NCs. As shown in Figure 6c, the absorption peaks in the UV-vis absorption spectra were detected at 396.5 nm, 375.0 nm, and 356.0 nm, respectively. The absorption peak of 396.5 nm is attributed to the transition from Br(4p) to Pb(6p) orbitals, whereas the 375.0 nm corresponds to the transition from Pb(6s) to the Pb(6p) orbitals [46]. The PL spectra are composed of multiple peaks in the deep blue spectrum region, and Figure 6d shows the multiple fitted PL data. The peaks are at 470.5, 459.1, 432.4, and 409.2 nm, respectively. The broadband emission and multiple-peak emission are attributed to the polydispersity of QD's spatial size (see Figure 5b) [47]. The Stokes shift and the FWHM of peak 4 are 12.7 nm and 17.0 nm, respectively. This small value of Stokes shift and FWHM implies that the peak of 4 results from the band-edge radiation in the form of free excitons [46]. However, the other peaks have larger Stokes shifts (more than 35.9 nm) and FWHMs (more than 24.6 nm). This phenomenon may take place due to the electron-phonon coupling, which is responsible for the increment of Stokes shift.
Photonics 2023, 10, x FOR PEER REVIEW 7 of 14 Figure 5 shows the size distribution of CsPbBr3 QDs based on the aforementioned HR-TEM images. The pristine CsPbBr3 QDs exhibit an average QD size of ~22 nm. However, the range is somewhat broad, from ~7 nm to ~50 nm, indicating the polydispersity of nanoparticles in this SR method at room temperature, although the chosen conditions (OA and OAm ratio) should affect this QD size and distribution. Importantly, when ZnBr2 was employed as a dopant for CsPbBr3 QDs, the QD size is in the range of ~1.5 to ~5.5 nm, with the average size of ~3 nm, guaranteeing the quantum size and confinement effect, a blue-light emission in this study. The optical properties of CsPbBr3 without and with ZnBr2 doping are summarized in Figure 6. As indicated above, the size of ZnBr2-doped CsPbBr3 QDs was about 3 nm, which is much less than the Bohr exciton radius of ~7 nm [14]. The very small size of the quantum dot was responsible for the significant blue shift of absorption and emission peaks ( Figure  6c) due to the quantum-confinement effect. Moreover, the blue shift might be assisted by the lattice internal stress, owing to the zinc-ion doping in the perovskite NCs. As shown in Figure 6c, the absorption peaks in the UV-vis absorption spectra were detected at 396.5 nm, 375.0 nm, and 356.0 nm, respectively. The absorption peak of 396.5 nm is attributed to the transition from Br(4p) to Pb(6p) orbitals, whereas the 375.0 nm corresponds to the transition from Pb(6s) to the Pb(6p) orbitals [46]. The PL spectra are composed of multiple peaks in the deep blue spectrum region, and Figure 6d shows the multiple fitted PL data. The peaks are at 470.5, 459.1, 432.4, and 409.2 nm, respectively. The broadband emission and multiple-peak emission are attributed to the polydispersity of QD's spatial size (see Figure 5b) [47]. The Stokes shift and the FWHM of peak 4 are 12.7 nm and 17.0 nm, respectively. This small value of Stokes shift and FWHM implies that the peak of 4 results from the band-edge radiation in the form of free excitons [46]. However, the other peaks have larger Stokes shifts (more than 35.9 nm) and FWHMs (more than 24.6 nm). This phenomenon may take place due to the electron-phonon coupling, which is responsible for the increment of Stokes shift.  Figure 7 shows the determination of optical band gaps of CsPbBr 3 QDs and ZnBr 2doped CsPbBr 3 QDs based on the Tauc model. From the UV−Vis absorption spectra, the bandgap was quantified by extrapolating the straight-line portion of the Tauc plot of (αhν) 2 vs. hν, in which h is Plank's constant, ν is frequency of incident photons, and α is absorption coefficient [48]. Resultantly, the bandgap is 2.30 eV for the pure CsPbBr 3 and 3.02 eV for ZnBr 2 -doped CsPbBr 3 , respectively.
To demonstrate the defect densities and size effects of the CsPbBr 3 QDs, the TRPL decay spectra were analyzed. The PL decay curve, as shown in Figure 8, can be well fitted with a bi-exponential function Equation (1) [49].
where A 0 represents a constant and A 1 and A 2 are the weights of multiple exponential functions constants, whereas τ 1 and τ 2 indicate short and long lifetimes originating from the excitonic radiative recombination and the trap-assisted nonradiative recombination, respectively. The average lifetime (τ ave ) was calculated using Equation (2) and the fitting parameters are shown in Table 1. Figure 8 and Table 1 show that the lifetime is shorter for smaller-size QDs (ZnBr 2 -doped CsPbBr 3 QDs). A longer lifetime is due to longer exciton (charge) diffusion length before its recombination (radiative and nonradiative). It is rational for the smaller-size QDs to have a shorter lifetime [32,50,51] because smallersize QDs have a higher surface area-to-volume ratio, resulting in high defect density and faster charge recombination. Here, it is noteworthy that any crystals with grain and grain boundaries (including small nanocrystals) are metastable (or unstable) compared to the bulk single crystals, because the surface energy is high enough, indicating that the phase transformation could be undergone by lowering energy to reach the equilibrium state, although the kinetics is unknown.
Photonics 2023, 10, x FOR PEER REVIEW 8 of 14  absorption coefficient [48]. Resultantly, the bandgap is 2.30 eV for the pure CsPbBr3 and 3.02 eV for ZnBr2-doped CsPbBr3, respectively. To demonstrate the defect densities and size effects of the CsPbBr3 QDs, the TRPL decay spectra were analyzed. The PL decay curve, as shown in Figure 8, can be well fitted with a bi-exponential function Equation (1) [49].
where 0 A represents a constant and 1 A and 2 A are the weights of multiple exponential functions constants, whereas 1 τ and 2 τ indicate short and long lifetimes originating from the excitonic radiative recombination and the trap-assisted nonradiative recombination, respectively. The average lifetime ( ave τ ) was calculated using Equation (2) and the fitting parameters are shown in Table 1. Figure 8 and Table 1 show that the lifetime is shorter for smaller-size QDs (ZnBr2-doped CsPbBr3 QDs). A longer lifetime is due to longer exciton (charge) diffusion length before its recombination (radiative and nonradiative). It is rational for the smaller-size QDs to have a shorter lifetime [32,50,51] because smaller-size QDs have a higher surface area-to-volume ratio, resulting in high defect density and faster charge recombination. Here, it is noteworthy that any crystals with grain and grain boundaries (including small nanocrystals) are metastable (or unstable) compared to the bulk single crystals, because the surface energy is high enough, indicating that the phase transformation could be undergone by lowering energy to reach the equilibrium state, although the kinetics is unknown.  At this moment, it is notable that when we added ZnBr2 into the CsPbBr3 perovskite precursor solution, the doping effect and the QD size reduction occur simultaneously. This fact indicates that it is very hard to examine a plain zinc-ion doping effect because the particle size changes simultaneously. Hence, we would like to investigate this kind of  At this moment, it is notable that when we added ZnBr 2 into the CsPbBr 3 perovskite precursor solution, the doping effect and the QD size reduction occur simultaneously. This fact indicates that it is very hard to examine a plain zinc-ion doping effect because the particle size changes simultaneously. Hence, we would like to investigate this kind of doping effect through the theoretical calculation based on the 3 × 3 × 3 supercell-sized CsPbBr 3 materials as shown in Figure 9a,b. To see the inherent electronic properties of Zn 2+ -doped CsPbBr 3 , we have calculated the electronic band structure and density of states. The PBE functional band structure of undoped and zinc-ion-doped 3 × 3 × 3 CsPbBr 3 supercells are shown in Figure 9c,d, respectively. However, Ghaithan et al. noted that the PBE pseudopotentials underestimate the bandgap of lead halide perovskites [52]. In our calculation, the bandgap of undoped 3 × 3 × 3 CsPbBr 3 supercells is about 2.01 eV (Figure 9c), whereas that of the doped sample was 1.72 eV (Figure 9d). Moreover, the bandgap analysis using the total density of states (DOS) shown in Figure 10a,b indicates that the bandgap decreased for the Zn 2+ -doped CsPbBr 3 . This trend of the bandgap reduction, when Pb 2+ was replaced by Zn 2+ in CsPbBr 3 , is in line with Guo et al.'s results [53]. However, in the case of our experimental results, the simultaneous change in QD size upon ZnBr 2 doping makes it very difficult to examine a pure doping effect. Hence, through the size reduction from~22 nm (without Zn 2+ ) to~3 nm (Zn 2+ doped), the doped sample exhibited a wider bandgap (i.e., a blue emitter) than the non-doped sample because of the apparent quantum size and confinement effects.

Conclusions
This study demonstrates the successful synthesis of strongly quantum-confined ZnBr 2doped CsPbBr 3 QDs under ambient conditions using the supersaturated recrystallization method. In this method, controlled amounts of ZnBr 2 were incorporated into the reaction medium to control the photophysical properties of the synthesized CsPbBr 3 QDs. As a result, 3 nm-sized blue-light-emitting CsPbBr 3 QDs were synthesized, covering the broad blue spectrum. The blue shifting of the emission spectrum is mainly attributed to the size reduction of the QDs. Our findings provide new insight into the synthesis of blue light-emitting CsPbBr 3 QDs at room temperature under ambient conditions. Future work may include the application of CsPbBr 3 QDs to photonic and optoelectronic devices including biosensors. Funding: This research received no external funding. The APC was funded by J.Y.K.

Data Availability Statement:
The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.