Exploring optimal in situ fabrication conditions to realize core–shell CsPbBr₃ QDs with high PLQYs and structural stability by dual-defect passivation

Abstract

Core–shell CsPbBr₃ QDs (core–shell M–CsPbBr₃ QDs) with high structural stability and excellent optical properties were developed through dual-defect passivation, with simultaneous application of in situ thiol ligand passivation and a core–shell structure using SiO₂ as a shell. When MPTES was injected immediately before Cs-oleate injection, the formation of by-products (PbS and trigonal-Cs₄PbBr₆ nanocrystals) was suppressed/minimized and effective surface defect passivation was achieved, resulting in defect-less core–shell M–CsPbBr₃ QDs. The thiol group of MPTES effectively passivated uncoordinated Pb²⁺ defects, while the SiO₂ shell formed by the hydrolysis reaction of three silyl ethers inhibited the formation of defects (vacancies) by preventing the penetration of moisture. The core–shell M–CsPbBr₃ QDs exhibited a PLQY of ∼82.9 ± 3.8%, much higher than that of pristine CsPbBr₃ QDs (∼65.3 ± 3.8%). Furthermore, they also showed more than 5 times higher structural stability in DI water compared to pristine CsPbBr₃ QDs. These results demonstrated that the synergistic effect of surface passivation with the thiol group and the core–shell structure can significantly improve the PLQYs and structural stability of CsPbBr₃ QDs.

---

Introduction

Recently, there has been intensive research on halide perovskite quantum dots (PQDs) due to their excellent optoelectronic properties.^1,2 In particular, CsPbX₃ QDs (X = Cl, Br, and I) have shown great potential in the fields of various optoelectronics including light-emitting diodes (LEDs),³ bio-imaging,⁴ and more.⁵ Nevertheless, CsPbX₃ QDs suffer from their poor structural stability to various external stimuli, especially moisture (water), due to their ionic nature. The surface defects and vacancies generated by ligand dissociation and halide ion migration cause rapid structural degradation of the CsPbX₃ QDs.^6,7 This structural degradation increases non-radiative recombination and consequently reduces the optoelectronic properties, including photoluminescence quantum yields (PLQYs).^8–10 Therefore, the development of methods for improving the structural stability of CsPbX₃ QDs by suppressing defect formation and passivating surface defects is urgently needed for their widespread practical applications. Various methods have been proposed to improve the structural stability of CsPbX₃ QDs, including metal (alkali metals, transition metals, etc.) doping,^11–13 core–shell structures,^14–16 and surface defect passivation such as ligand modification.^17,18 Metal doping is a method in which metal ions with the same charge number and similar ionic radii occupy the A-, B-, or interstitial sites of CsPbX₃ QDs.^19–21 The decrease of defect density and the change in the B–X bond length lead to an improvement in the structural stability of CsPbX₃ QDs.^11,22

The core–shell structure also enhanced the structural stability of CsPbX₃ QDs by introducing polymers or inorganic materials, especially oxide materials such as SiO₂, as shell layers.^5,23 The SiO₂ shell layer provided environmental and chemical resistance against external stimuli (e.g. moisture, oxygen, heat, etc.) and prevented the anion exchange between CsPbX₃ QDs with different halide compositions, enabling full-color displays through the stacking of red, green, and blue CsPbX₃ QDs.²⁴ The SiO₂ shell could be formed either through post-treatment^25–27 or an in situ process during the fabrication of CsPbX₃ QDs.^23,28 Core–shell CsPbBr₃/SiO₂ QDs were fabricated by post-treatment with tetraethyl orthosilicate (TEOS)²⁴via a hydrolysis reaction of the silane group of TEOS in air. Despite significant improvement in structural stability, the aggregation of CsPbBr₃/SiO₂ QDs and the presence of multiple CsPbBr₃ QDs within a single SiO₂ shell should be addressed for further perovskite LED (PeLED) applications.²⁹In situ SiO₂ shell formation can address the disadvantages of the post-treatment process mentioned above. Recently, our group successfully reported the fabrication of core–shell CsPbBr₃/SiO₂ QDs via an in situ hot-injection process.²³ 3-aminopropyl-triethoxysilane (APTES), which served not only as an amine ligand but also as a precursor for SiO₂ shell formation, was added during the fabrication of CsPbBr₃ QDs using a modified hot-injection process. The fabricated core–shell CsPbBr₃/SiO₂ QDs were agglomeration-free and consisted of a single CsPbBr₃ QD coated with an SiO₂ shell. They also showed excellent chemical stability and high PLQYs in various polar solvents. Notably, the thickness of a SiO₂ shell is ∼2–3 nm, which allowed core–shell CsPbBr₃/SiO₂ QDs to be applied to PeLEDs without compromising charge mobility.

Surface defect passivation using ligands with various functional groups (e.g., conjugated ligands or X-type ligands) is widely utilized to improve the structural stability of CsPbX₃.^30,31 Due to the low migration energy of halide ions, defects (vacancies) are easily formed, which destroy the PbX₆ octahedral structure and create uncoordinated Pb²⁺ defects.^32,33 Therefore, effective passivation of both halide defects and uncoordinated Pb²⁺ defects is strongly required to improve the structural stability of CsPbX₃ QDs. The introduction of ligands with thiol (–SH) groups as surface passivation ligands can significantly improve the structural stability and optoelectronic properties of CsPbBr₃ QDs.^34–38 The thiol group loses H⁺ to become S⁻, and S⁻ strongly binds to uncoordinated Pb²⁺ defects, effectively passivating (occupying) the halide defects and consequently healing the B- and X-site defects.^34,38 Surface defect passivation using ligands with thiol groups has been performed either as a post-treatment^34,39 or an in situ process.^36–38 Bi et al. reported that surface defects generated by ligand dissociation during the purification of CsPbI₃ QDs were successfully passivated by ligand exchange with the 2-aminoethanethiol (AET) ligand via the post-treatment process.³⁹ In most of the reported studies, the thiol ligand exchange, which is an X-type ligand, has been performed as a post-treatment process after the fabrication of PQDs because the in situ addition of thiol ligands instead of oleic acid (OA) and oleylamine (OAm) during the fabrication of PQDs can induce size/shape changes³⁴ and phase transitions^35,36,40 due to their strong affinity for Pb²⁺.

Recently, an in situ passivation strategy using various thiol ligands has been reported.^36,37 Yang et al. reported in situ surface defect passivation of CsPbBr₃ QDs with dodecylbenzenesulfonic acid.³⁸ This metal chelating sulfonate ligand has a strong affinity for Pb²⁺, which significantly enhanced the structural stability and PLQYs compared to those of CsPbBr₃ QDs passivated with OA and OAm.³⁸ This in situ passivation with a thiol ligand has the following advantages over post-treatment processes: (1) uniform particle size distribution via the fabrication of thermodynamically stable CsPbBr₃ QDs using a high-affinity thiol ligand, (2) suppression of surface defect formation through strong Pb–S binding, and (3) a simple fabrication protocol. L. Ruan et al. fabricated CsPb₂Br₅ nanowires (NWs) and nanosheets (NSs) with high PLQYs by in situ passivation using an alkyl thiol as the surface ligand during the fabrication of NWs and NSs via a precipitation method.³⁶ Nevertheless, the most serious problem associated with in situ thiol ligand passivation is the easy formation of PbS nanocrystals instead of halide perovskite nanocrystals due to the strong affinity of the sulfur group for Pb²⁺.^35–37 It also induces shape changes (e.g., from zero-dimensional QDs to one- or two-dimensional NWs and NSs) and phase transitions (e.g., from orthorhombic-CsPbBr₃ to tetragonal-CsPb₂Br₅ or trigonal-Cs₄PbBr₆).^35–37 Therefore, an optimal synthetic protocol must be developed to perform an in situ thiol ligand passivation method that can effectively passivate surface defects in CsPbBr₃ QDs while simultaneously suppressing PbS nanocrystal formation.

In this work, we have developed core–shell CsPbBr₃ QDs with high structural stability and PLQYs through dual-defect passivation, with simultaneous application of in situ thiol ligand passivation and a core–shell structure using SiO₂ as a shell. Most of the studies reported to date have enhanced the structural stability and optoelectronic properties of CsPbBr₃ QDs by suppressing surface defect formation through in situ thiol ligand passivation and the use of a core–shell structure. To the best of our knowledge, no studies combining these two methods have been reported. The synergistic effect of two representative defect passivation/suppression methods will dramatically improve the structural stability and optoelectronic properties of CsPbBr₃ QDs. (3-mercaptopropyl)triethoxysilane (MPTES) having a thiol group (–SH) and three silyl ether groups (–Si(OC₂H₅)₃) was used as a surface passivation ligand as well as a precursor for SiO₂ shell formation for dual-defect passivation (Fig. S1†). The thiol group of MPTES effectively passivated uncoordinated Pb²⁺ defects, while the SiO₂ shell formed by the hydrolysis reaction of three silyl ethers inhibited the formation of surface defects (vacancies) by preventing the penetration of moisture (water). Notably, in this work, dual-defect passivation is performed by an in situ process, rather than a post-treatment process. The in situ thiol ligand passivation of core–shell CsPbBr₃ QDs (i.e., core–shell M–CsPbBr₃ QDs) and the effect of dual-defect passivation on the structural stability and PLQYs of core–shell M–CsPbBr₃ QDs have been investigated. In particular, the optimal fabrication conditions for core–shell M–CsPbBr₃ QDs have been investigated by controlling synthetic parameters such as the MPTES injection point, the MPTES concentration (MPTES/Pb mol ratio), and the reaction time to suppress the formation of PbS and Cs₄PbBr₆ nanocrystals that degraded the optoelectronic properties. The crystal growth of core–shell M–CsPbBr₃ QDs was governed by both thiol (MPTES) and carboxylic acid (OA) ligands when the MPTES/Pb mol ratio was less than 1.0, but was mainly dominated by the thiol ligand when the MPTES/Pb mol ratio was more than 1.0. Therefore, when MPTES with a high binding affinity for Pb²⁺ was used as a ligand, a long reaction time (∼20 s) was required to grow core–shell M–CsPbBr₃ QDs thermodynamically rather than kinetically (∼5 s). When a high concentration of MPTES was injected after immediately degassing the Pb-pot, PbS, and trigonal-Cs₄PbBr₆ nanocrystals were fabricated instead of core–shell M–CsPbBr₃ QDs due to the strong affinity between Pb²⁺ and the thiol ligand and depletion of Pb²⁺. However, when a high concentration of MPTES was injected immediately before Cs-oleate injection, the formation of by-products (PbS and trigonal-Cs₄PbBr₆ nanocrystals) was suppressed/minimized and effective surface defect passivation was achieved, resulting in defect-less core–shell M–CsPbBr₃ QDs. The core–shell M–CsPbBr₃ QDs exhibited excellent PLQYs and structural stability compared to pristine CsPbBr₃ QDs. The core–shell M–CsPbBr₃ QDs exhibited a PLQY of ∼82.9 ± 3.8%, much higher than that of pristine CsPbBr₃ QDs (∼65.3 ± 3.8%). Furthermore, they also showed superior structural stability after exposure to deionized (DI) water, retaining about half of the initial PLQY (∼46.3%) after 9 h, while pristine CsPbBr₃ QDs were completely quenched after 5 h. Core–shell M–CsPbBr₃ QDs were found to be about 5 times more stable in DI water than pristine CsPbBr₃ QDs. These results demonstrated that dual-defect passivation with MPTES significantly improved the PLQYs and structural stability of CsPbBr₃ QDs, which can be applied to various optoelectronics in the future.

Experimental methods

Materials

Cesium carbonate (Cs₂CO₃, 99%), lead bromide (PbBr₂, 98%), oleic acid (OA, 90%), oleylamine (OAm, 70%), (3-mercaptopropyl)triethoxysilane (MPTES, 80%), 1-octadecane (ODE, 90%), and toluene (99.8%) were purchased from Sigma-Aldrich. Methyl acetate (MeAc, 99%) and acetone (99.7%) were purchased from Samchun. All chemicals were used without further purification.

Fabrication of pristine CsPbBr₃ QDs

For the preparation of the cesium oleate (Cs-oleate) precursor (Cs-pot), 100 mg of Cs₂CO₃ (0.614 mmol), 0.4 mL of OA, and 5 mL of ODE were added to a 25 mL two-neck round flask and degassed at ∼100 °C with continuous stirring under vacuum until obtaining a clear solution (∼30 min). In another 25 mL two-neck round flask (Pb-pot), 69 mg of PbBr₂ (0.188 mmol), 0.5 mL of OA and 0.5 mL of OAm were dissolved in 5 mL of ODE and degassed at ∼100 °C with continuous stirring under vacuum until obtaining a clear solution (∼30 min). The Cs-pot was heated to ∼150 °C and the Pb-pot was heated to ∼180 °C with continuous stirring under N₂ gas. Then, 0.4 mL of Cs-oleate was quickly injected into the Pb-pot (∼180 °C). After the reaction for ∼5 s, the mixture was cooled with ice water (crude solution). The crude solution was centrifuged with MeAc (crude solution : MeAc = 2 : 1, volume ratio) at 15 krpm for 5 min for purification. The precipitate was then dispersed in 2 mL of toluene and centrifuged again at 15 krpm for 5 min. The supernatant was placed in a brown vial and stored in an N₂-filled glovebox.

Fabrication of core–shell M–CsPbBr₃ QDs

Core–shell M–CsPbBr₃ QDs were fabricated via the same synthetic procedure as pristine CsPbBr₃ QDs except for the following conditions: (1) addition of a high-affinity thiol ligand, MPTES, (2) a different reaction time and (3) a different purification solvent (acetone). MPTES was injected into the Pb-pot at MPTES/Pb mol ratios ranging from 0.3 to 3.0. MPTES was injected into the Pb-pot at two different times: (1) immediately after degassing the Pb-pot and (2) immediately before Cs-oleate injection. The crude solution was prepared with a reaction time of ∼5–∼30 s and centrifuged at 15 krpm for 5 min with acetone (crude solution : acetone = 4 : 1, volume ratio) for ultra-thin SiO₂ shell formation by a hydrolysis reaction. The crude solution was centrifuged with MeAc (crude solution : MeAc = 2 : 1, volume ratio) at 15 krpm for 5 min for purification. The precipitate was then dispersed in 2 mL of toluene and centrifuged again at 15 krpm for 5 min. The supernatant was placed in a brown vial and stored in an N₂-filled glovebox.

Fabrication of MPTES–CsPbBr₃ QDs

MPTES–CsPbBr₃ QDs were fabricated via the same synthetic procedure as that of pristine CsPbBr₃ QDs, except for substituting OA with MPTES (OA-free). After degassing the Pb-pot, MPTES (MPTES/Pb mol ratio of 1.0) was injected into the OA-free Pb-pot. The crude solution was prepared with a reaction time of ∼5 s and centrifuged with MeAc (crude solution : MeAc = 2 : 1, volume ratio) at 15 krpm for 5 min for purification. The precipitate was then dispersed in 2 mL of toluene and centrifuged again at 15 krpm for 5 min. The supernatant was placed in a brown vial and stored in an N₂-filled glovebox.

Preparation of PbS nanocrystals

After degassing the Pb-pot, MPTES (MPTES/Pb mol ratios of 0.6 and 1.4) was injected into the Pb-pot, slowly heated to ∼180 °C and cooled with ice water without Cs-oleate injection. The crude solution was centrifuged with acetone (crude solution : acetone = 4 : 1, volume ratio) at 15 krpm for 5 min for purification. The precipitate was then dispersed in 2 mL of toluene and centrifuged again at 15 krpm for 5 min. The supernatant was placed in a brown vial and stored in an N₂-filled glovebox.

Characterization

The photoluminescence (PL) spectra were recorded using a PL spectrometer system consisting of two monochromators (SP-2150i and SP-2300i, Acton), a photomultiplier tube (PMT, Acton PD471) and a xenon lamp as an excitation source. The photoluminescence quantum yields (PLQYs) were calculated relative to coumarin 500 with a PLQY of ∼47% in ethanol under 390 nm excitation.⁴¹ Absorbance spectra were recorded using a double-beam UV-Vis spectrophotometer (Mega-800, Scinco). Raman spectra were recorded using a confocal Raman system equipped with a 785 nm excitation laser (I0785SD0100B-IS, Innovative Photonic Solutions), a 100× objective lens (NA 0.9, MPLFLN, Olympus), a Czerny–Turner spectrograph (Omni-λ200i, Zolix), and a low dark current deep-depletion charge-coupled device (iVac 316, Andor Technology). The crystal structures of the PQDs were obtained using a multi-purpose X-ray diffraction (XRD) system (SmartLab, Rigaku) with Cu Kα radiation (λ = 1.5418 Å). The chemical states of the PQDs were confirmed through X-ray photoelectron spectroscopy analysis (XPS, NEXSA, Thermo Fisher Scientific). The microstructures, particle size, and elemental composition of the PQDs were determined by transmission electron microscopy (TEM, Tecnai G2 F30 S-Twin 300 kV) and energy dispersive X-ray spectroscopy (EDS, Octane Elite EDS system). Fast Fourier transform (FFT) pattern analysis was performed using Gatan Digital Micrograph software (Gatan Inc.).

Results and discussion

The fabrication of in situ core–shell M–CsPbBr₃ QDs was optimized by controlling the following fabrication conditions: (1) MPTES injection time, (2) MPTES concentration (MPTES/Pb mol ratio), and (3) reaction time. Despite the high affinity of MPTES for Pb²⁺, the formation of PbS nanocrystals, which hindered the fabrication of core–shell M–CsPbBr₃ QDs, could be suppressed by controlling the fabrication conditions mentioned above. In particular, the effect of MPTES injection time on the optical properties of core–shell M–CsPbBr₃ QDs was systematically investigated. MPTES was injected into the Pb-pot at two different times during the fabrication of core–shell M–CsPbBr₃ QDs: (1) immediately after degassing the Pb-pot and (2) immediately before Cs-oleate injection. It is noteworthy that the MPTES concentration (MPTES/Pb mol ratio) and reaction time, which are dependent on the MPTES injection time, were also key parameters for the fabrication of high-quality core–shell M–CsPbBr₃ QDs via a modified in situ hot-injection method.

The effect of MPTES injection time and concentration (MPTES/Pb mol ratio) on the optical properties of core–shell M–CsPbBr₃ QDs was investigated. Fig. 1(a)–(c) show the PL spectra, PL_λmax, full width at half maximum (FWHM), and PLQYs of core–shell M–CsPbBr₃ QDs as a function of MPTES concentration (MPTES/Pb mol ratio). MPTES was injected into the Pb-pot immediately after degassing the Pb-pot. The PL spectral profiles (i.e., shape and PL_λmax) of core–shell M–CsPbBr₃ QDs varied around an MPTES/Pb mol ratio of 1.0 (Fig. 1(a)). Broad (FWHM: ∼30 nm) and non-Gaussian-shaped PL spectra centered at ∼500 nm were observed when the MPTES/Pb mol ratios were less than 1.0, whereas narrow (FWHM: ∼22 nm) and Gaussian-shaped PL spectra were observed when they were more than 1.0. The PL shoulder peak at ∼480 nm observed at MPTES/Pb mol ratios below 1.0 disappeared as MPTES/Pb mol ratios increased above 1.0. The PL spectra of core–shell M–CsPbBr₃ QDs were analyzed using the spectral deconvolution method, and the results are shown in Fig. S2.† The PL spectra could be resolved into two PL peaks (PL_λmax) at ∼503 nm and ∼509 nm regardless of MPTES/Pb mol ratios. These deconvolution results are consistent with the PL peak observed at ∼504 nm for MPTES–CsPbBr₃ QDs using MPTES (Pb-thiolate) as a ligand instead of OA and the PL peak observed at ∼510 nm for pristine CsPbBr₃ QDs using OA (Pb-oleate) as a ligand (Fig. S3†). It is well known that the Pb²⁺ release rate determines the shape, phase, and/or size of CsPbBr₃ QDs.³⁷ The strong affinity of the thiol group for Pb²⁺ slowly released Pb²⁺ from Pb-thiolate during the fabrication of the core–shell M–CsPbBr₃ QDs, causing a blue shift of the PL emission compared to pristine CsPbBr₃ QDs by inhibiting the crystal growth (i.e., small particle size). As the MPTES/Pb mol ratio increased, the deconvoluted PL intensity at ∼503 nm (red square) originating from core–shell M–CsPbBr₃ QDs fabricated from Pb-thiolate increased, while the deconvoluted PL intensity at ∼509 nm (blue circle) originating from pristine CsPbBr₃ QDs fabricated from Pb-oleate decreased and approached zero (Fig. S2(f)†). These results demonstrated that both thiol and carboxylic acid (OA) ligands could affect the crystal growth of CsPbBr₃ QDs, but the thiol ligand with a strong binding affinity for Pb²⁺ has more impact on the crystal growth. These different Pb²⁺ release rates of each ligand resulted in a broad PL spectrum (broadened FWHM and/or PL peak splitting). When the thiol ligand concentration is equal to or higher than the Pb²⁺ concentration, the PL spectrum becomes narrow and PL emission at ∼503 nm is dominant (Fig. 1(b) and Fig. S2†).

Fig. 1 (a) PL spectra, (b) PL_λmax and FWHM, (c) PLQYs, (d) XPS spectra, (e) XRD patterns and (f) absorbance spectra of core–shell M–CsPbBr₃ QDs as a function of MPTES concentration (MPTES/Pb mol ratio). JCPDS card no. 01-073-2478 of trigonal-Cs₄PbBr₆ nanocrystals. MPTES was injected into the Pb-pot immediately after degassing the Pb-pot. Reaction time was ∼5 s.

The PLQYs of core–shell M–CsPbBr₃ QDs as a function of MPTES concentration (MPTES/Pb mol ratio) are shown in Fig. 1(c). In general, the PQDs passivated with high-affinity ligands show high PLQYs and structural (chemical) stability.⁴² Therefore, the CsPbBr₃ QDs passivated with the thiol ligand are expected to show a high PLQY, because the thiol ligand has a stronger surface binding affinity than the carboxylic acid ligand. However, contrary to expectations, core–shell M–CsPbBr₃ QDs showed a much lower PLQY than pristine CsPbBr₃ QDs passivated with OA. The relatively low PLQYs at MPTES/Pb mol ratios below 1.0 were attributed to insufficient surface defect passivation due to the deficiency of the thiol ligand. In addition, the steric hindrance of MPTES reduced surface defect passivation by OA, which increased defect density and consequently reduced PLQYs.^19,43 On the other hand, although lower than that of pristine CsPbBr₃ QDs, the PLQYs of core–shell M–CsPbBr₃ QDs increased at MPTES/Pb mol ratios above 1.0, which was attributed to the passivation of uncoordinated Pb²⁺ defects with the S⁻ of the thiol ligand due to strong Pb–S binding at a high MPTES concentration (MPTES/Pb mol ratio). The peaks at ∼163.4 eV and ∼165.4 eV assigned to the bound thiol ligands (Pb–S binding)⁴⁴ and free thiol ligands³⁴ were observed in X-ray photoelectron spectroscopy (XPS) results (Fig. 1(d)). As the MPTES concentration (MPTES/Pb mol ratio) increased, the intensity at ∼163.4 eV, which is attributed to the bound thiol ligands, increased. These bound thiol ligands can also passivate Br⁻ defects (vacancies), resulting in an improved PLQY.^30,45 Nevertheless, the formation of non-luminescent trigonal-Cs₄PbBr₆ nanocrystals hindered further PLQY enhancement in core–shell M–CsPbBr₃ QDs at MPTES/Pb mol ratios above 1.0. The strong and selective affinity of the thiol group for Pb²⁺ led to the depletion of Pb²⁺ during the fabrication of CsPbBr₃ QDs, resulting in the fabrication of trigonal-Cs₄PbBr₆ nanocrystals.³⁷ The presence of trigonal-Cs₄PbBr₆ nanocrystals was confirmed by the X-ray diffraction (XRD) pattern and the absorbance spectra (Fig. 1(e) and (f)). The peaks at ∼12.6° and ∼12.9° assigned to the (012) and (110) planes of trigonal-Cs₄PbBr₆ nanocrystals (JCPDS no. 01-073-2478)³⁷ and the peak at ∼313 nm (ref. 35) were observed, respectively.

Notably, the concentration of trigonal-Cs₄PbBr₆ nanocrystals increased with increasing MPTES concentration (MPTES/Pb mol ratio) (Fig. 1(e) and (f)), but the crystal structure of the core–shell M–CsPbBr₃ QDs remained an orthorhombic structure regardless of the concentration of MPTES (Fig. S4†). These results demonstrated that the depletion of Pb²⁺ during the fabrication of core–shell M–CsPbBr₃ QDs is affected by the MPTES concentration (MPTES/Pb mol ratio), but could potentially be suppressed by controlling the reaction time of Pb²⁺ and S⁻ by changing the MPTES injection time. Therefore, to suppress the formation of non-luminescent trigonal-Cs₄PbBr₆ nanocrystals, the optimal MPTES injection time at a high thiol ligand concentration should be investigated to minimize the depletion of Pb²⁺.

As discussed earlier, the strong affinity of the thiol group for Pb²⁺ readily formed PbS nanocrystals, which prevented the Pb²⁺ precursor from being used for the formation of CsPbBr₃ QDs. In particular, PbS nanocrystals formed more efficiently at high concentrations of the thiol ligand due to high S⁻ (sulfur) concentration. The formation of PbS nanocrystals must be suppressed because it not only interrupted the fabrication of CsPbBr₃ QDs but also degraded the optical properties and structural stability of CsPbBr₃ QDs. Therefore, it is necessary to investigate in detail why and when the formation of PbS nanocrystals and the depletion of Pb²⁺ occur. Gurin et al. reported that when the PbS nanocrystals grow from Pb²⁺ and thiols in the solution, the color of the solution changes from yellow to light brown or black.⁴⁶ When MPTES was injected immediately after degassing the Pb-pot, the color of the Pb-pot changed from yellow to brown or dark brown, indicating the formation of PbS nanocrystals. The color of the Pb-pot became darker (brown to dark brown) as the MPTES concentration (MPTES/Pb mol ratio) increased (Movies S1 and S2†). The transparent solution in the Pb-pot slowly changed when MPTES with a ratio of 0.6 to Pb²⁺ (Movie S1†) was injected, whereas it turned dark brown immediately after MPTES with a ratio of 1.4 to Pb²⁺ (Movie S2†) was injected.

The formation of PbS nanocrystals was also confirmed by XPS, XRD, and UV-Vis analysis results. The intensity of the Pb–S binding (bound thiol ligands) peak at ∼163.4 eV increased with increasing MPTES concentration (MPTES/Pb mol ratio) in the absence of Cs-oleate injection (Fig. 2(a)). Since the Pb–S binding peak at ∼163.4 eV came from both the passivation of uncoordinated Pb²⁺ defects (positive effect) and PbS nanocrystals (negative effect), the origin of the Pb–S peak needs to be confirmed by other analytical methods. The XRD pattern and UV-Vis absorbance showed that PbS nanocrystals formed with increasing MPTES concentration (MPTES/Pb mol ratio) (Fig. 2(b) and (c)). At a high MPTES concentration (MPTES/Pb mol ratio of 1.4), diffraction peaks corresponding to the (111) and (200) lattice planes at ∼26.0° and ∼30.1° were observed (JCPDS no. 00-005-0592), whereas at a low MPTES concentration (MPTES/Pb mol ratio of 0.6), no diffraction peaks were observed for PbS nanocrystals. Furthermore, the absorbance peak at ∼945 nm demonstrated the presence of PbS nanocrystals at a high MPTES concentration (MPTES/Pb mol ratio).⁴⁷ These results suggested that the Pb–S binding peak at ∼163.4 eV (Fig. 2(a)) originated from both PbS nanocrystals and the passivation of uncoordinated Pb²⁺ defects, but the contribution of PbS nanocrystals became larger as the MPTES concentration (MPTES/Pb mol ratio) increased. The presence of PbS nanocrystals induced depletion of Pb²⁺ in the Pb-pot during the fabrication of core–shell M–CsPbBr₃ QDs, leading to the formation of non-luminescent trigonal-Cs₄PbBr₆ nanocrystals.

Fig. 2 (a) XPS spectra, (b) XRD patterns, and (c) absorbance spectra of PbS nanocrystals in the absence of Cs-oleate injection fabricated at different MPTES concentrations (MPTES/Pb mol ratio). JCPDS card no. 00-005-0592 of PbS. MPTES was injected into the Pb-pot immediately after degassing the Pb-pot. Reaction time was ∼5 s.

To suppress/minimize the formation of undesirable by-products (PbS and trigonal-Cs₄PbBr₆ nanocrystals), a pre-mixed solution (MPTES and Cs-oleate) prepared by adding MPTES to the Cs-pot was injected into the Pb-pot. This method is expected to reduce the interaction between the thiol group and Pb²⁺. The Cs-oleate solution without MPTES was transparent, while the Cs-oleate solution with MPTES was slightly opaque at 150 °C (Fig. S5(a) and (b)†). This is speculated to be because Cs (an alkali metal), which acts as a catalyst, induces cross-linking of the silyl ether of MPTES to form a Si–O–Si matrix.^48–50 This is further confirmed by the following results. When the Cs-pot without MPTES was cooled to room temperature (RT), the Cs-oleate was precipitated as a salt. The precipitated Cs-oleate is insoluble in ODE at RT, so the solution can flow. However, when the Cs-pot with MPTES was cooled to RT, cross-linking and gelation occurred, resulting in significantly higher viscosity that impeded flow (Fig. S5(c) and (d)†). Therefore, it is difficult to expect the desired functionality from both thiol and silyl ether groups in the pre-mixed solution. In contrast to the optimal fabrication conditions, under which MPTES was added to the Pb-pot, the core–shell M–CsPbBr₃ QDs fabricated using the pre-mixed solution (MPTES and Cs-oleate) showed excessively cross-linked aggregates (Si–O–Si matrix), as shown in Fig. S5(e).† From these results, the use of the pre-mixed solution of MPTES and Cs-oleate is not considered as the appropriate method for the passivation of Pb²⁺ defects with the thiol group of MPTES. In addition, the thiol group reacts or binds with Cs⁺ in Cs-oleate solution, which slows down the formation of Pb-thiolate via a reaction with the Pb²⁺ precursor in the Pb-pot, resulting in poor Pb²⁺ defect passivation. Therefore, when a high concentration of the thiol ligand is used, the interaction between Pb²⁺ and the thiol ligand has to be reduced to minimize the formation of PbS nanocrystals, thus preventing depletion of Pb²⁺, which is responsible for the formation of non-luminescent trigonal-Cs₄PbBr₆ nanocrystals.

Fig. 3(a)–(c) show the XRD patterns and absorbance spectra of core–shell M–CsPbBr₃ QDs fabricated at different MPTES injection times. As mentioned above, when MPTES was injected into the Pb-pot after immediately degassing the Pb-pot, trigonal-Cs₄PbBr₆ nanocrystals were formed. However, the XRD peaks and absorbance peaks of trigonal-Cs₄PbBr₆ nanocrystals were barely observed when MPTES was injected into the Pb-pot immediately before Cs-oleate injection under the same fabrication conditions (MPTES/Pb mol ratio of 1.4 and ∼5 s reaction time). Furthermore, the color of the Pb-pot remained yellow until the core–shell M–CsPbBr₃ QDs were fabricated (Fig. 3(d)). These results demonstrated that no or little depletion of Pb²⁺ occurred during the fabrication of core–shell M–CsPbBr₃ QDs despite MPTES being used as a ligand. By optimizing the MPTES injection time (i.e., MPTES injection immediately before Cs-oleate injection) to reduce the interaction between Pb²⁺ and the thiol ligand, core–shell M–CsPbBr₃ QDs with high PLQYs can be fabricated by suppressing the formation of PbS and non-luminescent trigonal-Cs₄PbBr₆ nanocrystals and passivating uncoordinated Pb²⁺ defects.

Fig. 3 (a) XRD patterns, (b) magnified view in a range of 10° to 15°, and (c) absorbance spectra of core–shell M–CsPbBr₃ QDs with different MPTES injection times. Core–shell M–CsPbBr₃ QDs were fabricated under the same fabrication conditions (MPTES/Pb mol ratio of 1.4 and reaction time of ∼5 s) except for different MPTES injection times. The core–shell M–CsPbBr₃ QDs showed an orthorhombic crystal structure (JCPDS card no. 98-009-7851). (d) Digital photograph of the Pb-pot after MPTES injection. MPTES (MPTES/Pb mol ratio of 1.4) was injected into the Pb-pot immediately before Cs-oleate injection.

The optical properties of core–shell M–CsPbBr₃ QDs fabricated by injecting MPTES immediately before Cs-oleate injection were investigated. Fig. 4(a)–(c) show the PL properties (PL spectra, PL_λmax, FWHM, and PLQYs) of core–shell M–CsPbBr₃ QDs as a function of the reaction time under the fabrication conditions where the MPTES/Pb mol ratio was fixed at 1.4. Unlike pristine CsPbBr₃ QDs, the core–shell M–CsPbBr₃ QDs fabricated with a reaction time of ∼5 s showed a non-Gaussian-shaped PL spectrum consisting of two peaks at ∼480 nm and ∼504 nm (Fig. 4(a) and S6(a)†). The PL shoulder peak at ∼480 nm is speculated to be due to the formation of CsPbBr₃ QDs with a small particle size, which is commonly observed under unoptimized fabrication conditions (e.g., precursor ratio, reaction time, temperature, etc.).^23,51–53 The high affinity of the thiol group for Pb²⁺ caused slow nucleation and crystal growth during the fabrication of CsPbBr₃ QDs using MPTES as the ligand. The slow surface ligand dissociation rate of MPTES delayed the adsorption of the precursor, allowing core–shell M–CsPbBr₃ QDs to be grown thermodynamically rather than kinetically. The PL peak at ∼505 nm is direct evidence for core–shell M–CsPbBr₃ QDs fabricated using the thiol ligand, which is dominated by slow crystal growth (Fig. S6(b)†).

Fig. 4 (a) PL spectra, (b) PL_λmax and FWHM and (c) PLQYs of core–shell M–CsPbBr₃ QDs as a function of reaction time. The MPTES concentration (MPTES/Pb mol ratio) was fixed at 1.4. (d) PL spectra, (e) PL_λmax and FWHM and (f) PLQYs of core–shell M–CsPbBr₃ QDs as a function of the MPTES concentration (MPTES/Pb mol ratio). The reaction time was fixed at ∼20 s except for pristine CsPbBr₃ QDs (∼5 s). In both experiments, MPTES was injected into the Pb-pot immediately before Cs-oleate injection.

The absence of the Pb–S binding peak (∼163.4 eV) in the core–shell M–CsPbBr₃ QDs fabricated at a reaction time of ∼5 s demonstrated that the uncoordinated Pb²⁺ defects were not efficiently passivated by the bound thiol ligands (Fig. S7†). Consequently, the core–shell M–CsPbBr₃ QDs fabricated at a reaction time of ∼5 s resulted in a lower PLQY (∼40.4 ± 3.9%) compared to pristine CsPbBr₃ QDs (∼65.3 ± 3.8%) (Fig. 4(c)). The long reaction time (∼20 s) increased the binding probability between Pb²⁺ and the thiol ligand, which improved the optical properties of core–shell M–CsPbBr₃ QDs by passivating surface defects with thiol ligands (Fig. 4(c) and Fig. S7†). Furthermore, as the reaction time increased from ∼5 s to ∼20 s, the FWHM of core–shell M–CsPbBr₃ QDs also decreased from ∼28 nm to ∼20 nm (Fig. 4(b)). The uniform particle size of core–shell M–CsPbBr₃ QDs due to thermodynamic crystal growth is responsible for this narrow FWHM. In addition, the PL shoulder peak at ∼480 nm observed at a reaction time of ∼5 s completely disappeared at ∼20 s (Fig. S6†), resulting in a Gaussian-shaped PL spectrum. This suggested that the fabrication conditions for the core–shell M–CsPbBr₃ QDs were optimized by increasing the reaction time. As the reaction time increased, the PLQYs of core–shell M–CsPbBr₃ QDs gradually increased, reaching a maximum value of ∼64.7 ± 3.7% at a reaction time of ∼20 s, which is comparable to or slightly lower than that of pristine CsPbBr₃ QDs (65.3 ± 3.8%) (Fig. 4(c)). Note that the MPTES/Pb mol ratio was fixed at 1.4 in these experiments. The PLQYs of core–shell M–CsPbBr₃ QDs can be further improved by optimization of the MPTES/Pb mol ratio.

Fig. 4(d)–(f) show the PL properties (PL spectra, PL_λmax, FWHM, and PLQYs) of core–shell M–CsPbBr₃ QDs as a function of the MPTES concentration (MPTES/Pb mol ratio) under the fabrication conditions where the reaction time was fixed at ∼20 s. Unlike the core–shell M–CsPbBr₃ QDs fabricated with a reaction time of ∼5 s, no PL shoulder peak at ∼480 nm was observed for the core–shell M–CsPbBr₃ QDs fabricated with a reaction time of ∼20 s for all MPTES concentrations (MPTES/Pb mol ratios) (Fig. 4(d)). The PL_λmax and FWHM of all core–shell M–CsPbBr₃ QDs exhibited similar values of ∼506 nm (∼507 nm for an MPTES/Pb mol ratio of 3.0) and ∼20 nm, respectively (Fig. 4(e)). The narrow FWHM of core–shell M–CsPbBr₃ QDs (∼20 nm) enables excellent color purity in LED applications. The highest PLQY of 82.9 ± 3.8% was achieved at an MPTES/Pb mol ratio of 2.4. Under these fabrication conditions, the color of the Pb-pot did not turn brown and remained yellow until the core–shell M–CsPbBr₃ QDs were fabricated (Movie S3†), indicating that the formation of PbS nanocrystals did not occur, which is responsible for the formation of non-luminescent trigonal-Cs₄PbBr₆ nanocrystals by depletion of Pb²⁺. However, at an MPTES/Pb mol ratio of 3.0, the PLQY slightly decreased, which was attributed to the depletion of Pb²⁺ in the Pb-pot due to the high concentration of MPTES, as mentioned above. Fig. S8† shows that trigonal-Cs₄PbBr₆ nanocrystals precipitated in a crude solution with an MPTES/Pb mol ratio of 3.0 and a reaction time of ∼20 s.

Fig. S9† summarizes the effect of synthetic parameters (MPTES injection time, MPTES concentration (MPTES/Pb mol ratio), and reaction time) on the optical properties of core–shell M–CsPbBr₃ QDs. The crystal growth of core–shell M–CsPbBr₃ QDs was governed by both thiol (MPTES) and carboxylic acid (OA) ligands when the MPTES/Pb mol ratio was less than 1.0, but was mainly dominated by the thiol ligand when the MPTES/Pb mol ratio was more than 1.0. In general, CsPbBr₃ QDs kinetically grew when a carboxylic acid (OA) with a weak binding affinity for Pb²⁺ was used as a ligand. On the other hand, when MPTES, which has a high binding affinity for Pb²⁺, was used as a ligand, CsPbBr₃ QDs were more favorable to grow thermodynamically, thus requiring a longer reaction time (∼20 s) than kinetically grown QDs (∼5 s). The thermodynamic crystal growth suppressed the formation of defects, which enabled the implementation of high-quality CsPbBr₃ QDs. The MPTES injection time has a significant effect on the optical properties of core–shell M–CsPbBr₃ QDs. When a high concentration of MPTES (MPTES/Pb mol ratio was larger than 1.0) was immediately injected after degassing the Pb-pot, non-luminescent trigonal-Cs₄PbBr₆ nanocrystals were fabricated instead of core–shell M–CsPbBr₃ QDs, which was attributed to the depletion of Pb²⁺ by the formation of PbS nanocrystals due to the strong affinity between Pb²⁺ and the thiol ligand. However, when a high concentration of MPTES was injected immediately before Cs-oleate injection, the formation of by-products (PbS and trigonal-Cs₄PbBr₆ nanocrystals) was suppressed/minimized and effective surface defect passivation was achieved, resulting in defect-less core–shell M–CsPbBr₃ QDs. These results demonstrated that systematic control of synthetic parameters (concentration and injection time of the thiol ligand, reaction time, etc.) is necessary to ensure superior optical and structural stability of CsPbBr₃ QDs fabricated by a modified in situ hot-injection method using the thiol ligand rather than conventional OA.

The introduction of inorganic shells (e.g., SiO₂) into CsPbBr₃ QDs can significantly improve their structural stability against environmental stimuli (moisture, oxygen, light, etc.) without deterioration of the optoelectronic properties. MPTES not only acts as a ligand to passivate surface defects (Pb²⁺ and Br⁻ defects) but can also act as a precursor for SiO₂ shell formation. The thiol group (–SH) effectively passivated the surface defects, and three silyl ether groups formed a cross-linked Si–O–Si matrix coating each CsPbBr₃ QD through a hydrolysis reaction.^19,54 The silyl ether groups of MPTES reacted with polar media (e.g., acetone and methyl acetate) during the purification process to form a Si–O–Si matrix (SiO₂ shell), which was coated on the surfaces of CsPbBr₃ QDs. The chemical reactions for the formation of an SiO₂ shell are described below.

–SiOC₂H₅ + H₂O → –SiOH + C₂H₅OH (1)

–SiOH + –SiOC₂H₅ → –SiOSi– + C₂H₅OH (2)

–SiOH + –SiOH → –SiOSi– + H₂O. (3)

The high-resolution transmission electron microscopy (HR-TEM) image of core–shell M–CsPbBr₃ QDs in Fig. 5(c) confirmed the presence of an ultra-thin SiO₂ shell with a thickness of ∼1–2 nm, which was not observed in the HR-TEM image of the pristine CsPbBr₃ QDs (Fig. 5(a)). The transmission electron microscopy-energy dispersive spectroscopy (TEM-EDS) analysis further confirmed the presence of an SiO₂ shell on the surfaces of CsPbBr₃ QDs (Fig. S10†). The scanning transmission electron microscopy (STEM) image (Fig. S10(a)†) shows the location where TEM-EDS line scanning was performed to analyze the composition of core–shell M–CsPbBr₃ QDs. Since the long analysis time of TEM-EDS line scanning can lead to the degradation of core–shell M–CsPbBr₃ QDs, TEM-EDS line scanning was performed at low magnification rather than at high magnification to obtain reliable measurement results. Si and O were detected together in the region where the core–shell M–CsPbBr₃ QDs were clustered (Fig. S10(c)–(e) and Table S1†). From the TEM-EDS results, it is demonstrated that the SiO₂ shell was successfully coated on the surfaces of CsPbBr₃ QDs. The TEM results also showed that a single CsPbBr₃ QD was coated with an SiO₂ shell, rather than multiple CsPbBr₃ QDs embedded in an SiO₂ shell^22,53 and core–shell M–CsPbBr₃ QDs exhibited a well-ordered morphology without aggregation (Fig. S10(b)†).

Fig. 5 (a)–(c) HR-TEM images of (a) pristine CsPbBr₃ QDs, (b) shell-less M–CsPbBr₃ QDs, and (c) core–shell M–CsPbBr₃ QDs. MPTES was injected into the Pb-pot immediately before Cs-oleate injection. The MPTES concentration (MPTES/Pb mol ratio) was fixed at 2.4 and the reaction time was fixed at ∼20 s. The inset shows fast Fourier transform (FFT) images. The chemical/environmental stabilities of pristine CsPbBr₃ QDs, shell-less M–CsPbBr₃ QDs, and core–shell M–CsPbBr₃ QDs (d) immersed in DI water and (e) exposed to ambient air.

The Raman spectra also proved the presence of the SiO₂ shell (Fig. S11(a)†). A Raman peak at ∼1143 cm⁻¹, attributed to Si–O–Si stretching,⁵⁵ was observed in the core–shell M–CsPbBr₃ QDs, which was not observed in the pristine CsPbBr₃ QDs. In addition, the Raman peak at ∼1119 cm⁻¹ corresponding to the Si–O–C stretching⁵⁶ is attributed to the unhydrolyzed silyl ether groups of MPTES attached to the surfaces of core–shell M–CsPbBr₃ QDs. Shell-less M–CsPbBr₃ QDs were fabricated under the same fabrication conditions, except that only methyl acetate (MeAc) was used as a purification solvent. As shown in the HR-TEM image (Fig. 5(b)), a negligibly thin SiO₂ shell was formed on the surfaces of the CsPbBr₃ QDs. The Raman peak at ∼1119 cm⁻¹ (Si–O–Si) was observed, while the Raman peak at ∼1143 cm⁻¹ (Si–O–C) was weakly observed (Fig. S11(a)†). This is due to the relatively low polarity of MeAc compared to acetone, which makes the hydrolysis reaction less efficient. Furthermore, the XPS spectra showed that the intensity of the Si 2p (∼102.4 eV) peak increased gradually as the SiO₂ shell was formed (Fig. S11(b)†).

The d-spacing values of the three CsPbBr₃ QDs (pristine/shell-less/core–shell) were obtained from the fast Fourier transform (FFT) analysis of the HR-TEM images, and the results are shown in the insets of Fig. 5. The d-spacings of shell-less and core–shell M–CsPbBr₃ QDs are ∼2.93, ∼4.14, and ∼5.87 Å, corresponding to the (040), (121), and (020) planes, respectively, which are similar to those of pristine CsPbBr₃ QDs (∼2.95, ∼4.18, and ∼5.91 Å corresponding to the (040), (121), and (020) planes, respectively). In addition, the XRD patterns showed that the three CsPbBr₃ QDs exhibited an orthorhombic crystal structure (JCPDS no. 98-009-7851) (Fig. S11(c)†). These results suggested that no significant crystal structure transition occurred in CsPbBr₃ QDs after surface passivation by Pb–S binding and/or ultra-thin SiO₂ shell formation.

Fig. S12† shows the TEM images and the size distribution of the three CsPbBr₃ QDs. The three CsPbBr₃ QDs showed a uniform cube-like shape in the TEM images. The average particle sizes of the shell-less and core–shell M–CsPbBr₃ QDs were ∼9.64 and ∼9.71 nm, respectively, which were ∼1 nm smaller than that of the pristine CsPbBr₃ QDs (∼10.73 nm). The particle sizes of core–shell and shell-less M–CsPbBr₃ QDs were expressed simply as the size of the core (CsPbBr₃ QDs) without considering the SiO₂ shell thickness to ensure the accuracy of the particle size. It is consistent with the slight blue shift of the PL peaks (from ∼510 to ∼506 nm) in Fig. 4(e). The slightly smaller particle sizes of core–shell and shell-less M–CsPbBr₃ QDs are attributed to the slow crystal growth due to the strong affinity of the thiol group for Pb²⁺ compared to OA. The relative standard deviation (RSD) of the particle size distribution of shell-less and core–shell M–CsPbBr₃ QDs was reduced by about ∼4% compared to that of pristine CsPbBr₃ QDs, enabling high-purity displays with a narrow FWHM.

The chemical/environmental stabilities of the three CsPbBr₃ QDs (pristine/shell-less/core–shell) were investigated. Fig. 5(d) and (e) show the PL intensity variation of the three CsPbBr₃ QDs in response to external stimuli: (1) immersion in DI water (CsPbBr₃ QD solutions were mixed with 20 vol% DI water) and (2) exposure to ambient air. The PL intensity of the three CsPbBr₃ QDs gradually decreased as the time of exposure to DI water increased. However, the PL intensity decreased in the order of core–shell M–CsPbBr₃ QDs, shell-less M–CsPbBr₃ QDs, and pristine CsPbBr₃ QDs (Fig. 5(d) and Fig. S13†). The pristine CsPbBr₃ QDs maintained about half of the initial PL intensity (∼57.7%) after 1 hour and were completely quenched after 5 hours. Unlike pristine CsPbBr₃ QDs, the core–shell M–CsPbBr₃ QDs retained half of the initial PL intensity (∼46.3%) even after about 9 hours and were not completely quenched after even 14 hours. The PL quenching of shell-less M–CsPbBr₃ QDs proceeded at a rate intermediate between those of pristine CsPbBr₃ QDs and core–shell M–CsPbBr₃ QDs (∼43.3% reduction from initial PL intensity after about 3 h). The structural stability of core–shell M–CsPbBr₃ QDs in DI water was improved by at least 5 times compared to that of pristine CsPbBr₃ QDs. The PL quenching trends of the three CsPbBr₃ QDs upon air exposure were similar to those observed under DI water exposure (Fig. 5(e) and Fig. S14†). When pristine CsPbBr₃ QDs were exposed to ambient air, PL was almost quenched after 30 days. On the other hand, the structural stability of core–shell M–CsPbBr₃ QDs upon air exposure was significantly improved compared to that of pristine CsPbBr₃ QDs. When exposed to ambient air for 30 days, the PL intensity only decreased by ∼40% of the initial PL intensity. This result demonstrated that the thiol group and core–shell structure significantly enhanced the structural stability of CsPbBr₃ QDs against external stimuli (moisture, oxygen, light, etc.) through uncoordinated Pb²⁺ defect passivation and the prevention of moisture penetration. This result is consistent with previous studies, in which the environmental stability of PQDs was significantly improved by surface defect passivation with the thiol ligand and the introduction of an SiO₂ shell.^23,30,38,39 The work is expected to contribute significantly to future optoelectronic applications that require high structural stability and PLQYs, such as PeLEDs, bio-imaging, and gas sensors.

Conclusions

In this work, we have developed core–shell M–CsPbBr₃ QDs with high structural stability and excellent optical properties by simultaneously applying in situ thiol ligand passivation and a core–shell structure using SiO₂ as a shell. (3-mercaptopropyl)triethoxysilane (MPTES) with a thiol group (–SH) and three silyl ethers (–Si(OC₂H₅)₃) was used as a surface passivation ligand as well as a precursor for SiO₂ shell formation for dual-defect passivation. The thiol group of MPTES effectively passivated uncoordinated Pb²⁺ defects; however, unintentional by-products (PbS and trigonal-Cs₄PbBr₆ nanocrystals) were also fabricated instead of core–shell M–CsPbBr₃ QDs due to the strong affinity between Pb²⁺ and the thiol ligand. The formation of by-products was suppressed/minimized and effective surface defect passivation was achieved through systematic control of synthetic parameters (concentration and injection time of the thiol ligand, reaction time, etc.). The high affinity of the thiol group for Pb²⁺ slowly released Pb²⁺ from Pb-thiolate, causing slow crystal growth during the fabrication of core–shell M–CsPbBr₃ QDs. Therefore, a high concentration of MPTES and a long reaction time (∼20 s) are required to ensure that the crystal growth of core–shell M–CsPbBr₃ QDs is dominated by the thiol ligand and follows a thermodynamic rather than a kinetic reaction (∼5 s). The long reaction time (∼20 s) increased the binding probability between Pb²⁺ and the thiol ligand, which improved the optical properties of core–shell M–CsPbBr₃ QDs by passivating uncoordinated Pb²⁺ defects with thiol ligands. The injection time of MPTES should be delayed (immediately before Cs-oleate injection) to reduce interaction with Pb²⁺, thereby inhibiting the formation of PbS and trigonal-Cs₄PbBr₆ nanocrystals, which interrupt the formation of core–shell M–CsPbBr₃ QDs. The core–shell M–CsPbBr₃ QDs exhibited excellent optical properties and structural stability compared to pristine CsPbBr₃ QDs. The core–shell M–CsPbBr₃ QDs exhibited a high PLQY of ∼82.9 ± 3.8% and a narrow FWHM of ∼20 nm compared to pristine CsPbBr₃ QDs (∼65.3 ± 3.8% and ∼23 nm). Furthermore, they also showed higher structural stability against external stimuli (moisture, oxygen, light, etc.). In particular, core–shell M–CsPbBr₃ QDs retained approximately half of the initial PL intensity after 9 hours of immersion in DI water. These results demonstrated that dual-defect passivation with MPTES can significantly improve the PLQYs and structural stability of CsPbBr₃ QDs by passivating surface defects and preventing the penetration of moisture (water).

Data availability

The data supporting this article have been included as part of the ESI.†

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the Basic Science Research Program of the National Research Foundation of Korea (NRF) grants funded by the Korean government (MEST) (NRF-2022R1A2C1010773 and RS-2024-00357495).

References

1. Y. Zhou and Y. Zhao, Energy Environ. Sci., 2019, 12, 1495–1511.
2. K. K. Chan, D. Giovanni, H. He, T. C. Sum and K.-T. Yong, ACS Appl. Nano Mater., 2021, 4, 9022–9033.
3. H. C. Yoon, H. Kang, S. Lee, J. H. Oh, H. Yang and Y. R. Do, ACS Appl. Mater. Interfaces, 2016, 8, 18189–18200.
4. P. Kumar, M. Patel, C. Park, H. Han, B. Jeong, H. Kang, R. Patel, W.-G. Koh and C. Park, J. Mater. Chem. B, 2020, 8, 10337–10345.
5. C. Zhang, J. Chen, L. Kong, L. Wang, S. Wang, W. Chen, R. Mao, L. Turyanska, G. Jia and X. Yang, Adv. Funct. Mater., 2021, 31, 2100438.
6. G. Almeida, I. Infante and L. Manna, Science, 2019, 364, 833–834.
7. B. Rivkin, P. Fassl, Q. Sun, A. D. Taylor, Z. Chen and Y. Vaynzof, ACS Omega, 2018, 3, 10042–10047.
8. M. Z. Rahaman, S. Ge, C.-H. Lin, Y. Cui and T. Wu, Small Struct., 2021, 2, 2000062.
9. D. Huang, J. Bo, R. Zheng, C. Luo, X. Sun, Q. Li, D. Shen, Z. Zheng, M. Chen and Y. Yang, Adv. Opt. Mater., 2021, 9, 2100474.
10. C. Jo, D. Kim, C.-L. Lee and D.-K. Ko, Opt. Express, 2023, 31, 40352–40365.
11. S. Zou, Y. Liu, J. Li, C. Liu, R. Feng, F. Jiang, Y. Li, J. Song, H. Zeng and M. Hong, J. Am. Chem. Soc., 2017, 139, 11443–11450.
12. Y. Yang, Y. Li, W. Gong, H. Guo and X. Niu, Colloids Surf., A, 2023, 663, 131083.
13. B. Roy, S. Mahato, S. Bose, A. Ghorai, S. K. Srivastava, N. C. Das and S. K. Ray, Chem. Mater., 2023, 35, 1601–1609.
14. F. Zhang, Z. Shi, S. Li, Z. Ma, Y. Li, L. Wang, D. Wu, Y. Tian, G. Du and X. Li, ACS Appl. Mater. Interfaces, 2019, 11, 28013–28022.
15. H. Dong, H. Zhao, T. Xuan, W. Bai, T. Lin, Y. Cai and R.-J. Xie, ACS Appl. Mater. Interfaces, 2023, 15, 29297–29307.
16. Y. Cai, Y. Yang, H. Dong, T. Xuan, X. Tang and R.-J. Xie, J. Mater. Chem. C, 2023, 11, 9486–9494.
17. Y. Liu, X. Shao, Z. Gao, Q. Xie, Y. Ying, X. Zhu, Z. Pan, J. Yang, H. Lin and X. Tang, Small, 2024, 20, 2305664.
18. Z. Zeng, Y. Meng, Z. Yang, Y. Ye, Q. Lin, Z. Meng, H. Hong, S. Ye, Z. Cheng and Q. Lan, ACS Appl. Mater. Interfaces, 2024, 16(8), 10389–10397.
19. X. Tang, W. Chen, Z. Liu, J. Du, Z. Yao, Y. Huang, C. Chen, Z. Yang, T. Shi and W. Hu, Small, 2019, 15, 1900484.
20. M. T. Hoang, A. S. Pannu, C. Tang, Y. Yang, N. D. Pham, K. Gui, X. Wang, S. Yambem, P. Sonar and A. Du, Adv. Opt. Mater., 2020, 8, 2000742.
21. D. Kim, Y.-R. Jo, S. Kim, S.-Y. Yim, J.-H. Lee and C.-L. Lee, J. Mater. Chem. C, 2024, 12, 6395–6405.
22. A. Swarnkar, W. J. Mir and A. Nag, ACS Energy Lett., 2018, 3, 286–289.
23. C. K. Trinh, H. Lee, M. G. So and C.-L. Lee, ACS Appl. Mater. Interfaces, 2021, 13, 29798–29808.
24. F. Gao, W. Yang, X. Liu, Y. Li, W. Liu, H. Xu and Y. Liu, Chem. Eng. J., 2021, 407, 128001.
25. C. Collantes, V. G. Pedro, M.-J. Banuls and A. Maquieira, ACS Appl. Nano Mater., 2021, 4, 2011–2018.
26. C. Collantes, W. Teixeira, V. González-Pedro, M.-J. Bañuls, P. Quintero-Campos, S. Morais and Á. Maquieira, Dalton Trans., 2023, 52, 18464–18472.
27. C. Rossi, R. Scarfiello, R. Brescia, L. Goldoni, G. Caputo, L. Carbone, D. Colombara, L. De Trizio, L. Manna and D. Baranov, Chem. Mater., 2021, 34, 405–413.
28. W. Song, Y. Wang, B. Wang, Y. Yao, W. Wang, J. Wu, Q. Shen, W. Luo and Z. Zou, Nano Res., 2020, 13, 795–801.
29. Q. Zhong, M. Cao, H. Hu, D. Yang, M. Chen, P. Li, L. Wu and Q. Zhang, ACS Nano, 2018, 12, 8579–8587.
30. S. Park, H. Cho, W. Choi, H. Zou and D. Y. Jeon, Nanoscale Adv., 2019, 1, 2828–2834.
31. G. Li, J. Huang, H. Zhu, Y. Li, J.-X. Tang and Y. Jiang, Chem. Mater., 2018, 30, 6099–6107.
32. H. Wu, J. Qiu, J. Wang, Y. Wen, Q. Wang, Z. Long, D. Zhou, Y. Yang and D. Wang, Inorg. Chem. Front., 2021, 8, 658–668.
33. X. Chen, Z. Sun, B. Cai, X. Li, S. Zhang, D. Fu, Y. Zou, Z. Fan and H. Zeng, Adv. Funct. Mater., 2022, 32, 2112129.
34. L. Ruan, W. Shen, A. Wang, Q. Zhou, H. Zhang and Z. Deng, Nanoscale, 2017, 9, 7252–7259.
35. Z. Liu, Y. Bekenstein, X. Ye, S. C. Nguyen, J. Swabeck, D. Zhang, S.-T. Lee, P. Yang, W. Ma and A. P. Alivisatos, J. Am. Chem. Soc., 2017, 139, 5309–5312.
36. L. Ruan, W. Shen, A. Wang, A. Xiang and Z. Deng, J. Phys. Chem. Lett., 2017, 8, 3853–3860.
37. A. Dutta, R. K. Behera and N. Pradhan, ACS Energy Lett., 2019, 4, 926–932.
38. D. Yang, X. Li, W. Zhou, S. Zhang, C. Meng, Y. Wu, Y. Wang and H. Zeng, Adv. Mater., 2019, 31, 1900767.
39. C. Bi, S. V. Kershaw, A. L. Rogach and J. Tian, Adv. Funct. Mater., 2019, 29, 1902446.
40. S. K. Dutta, S. Bera, S. Sen and N. Pradhan, ACS Energy Lett., 2022, 7, 773–777.
41. S. Nad and H. Pal, J. Phys. Chem. A, 2003, 107, 501–507.
42. J. Ye, M. M. Byranvand, C. O. Martínez, R. L. Hoye, M. Saliba and L. Polavarapu, Angew. Chem., Int. Ed., 2021, 60, 21636–21660.
43. H. Lee, C. K. Trinh, M. G. So and C.-L. Lee, Nanoscale, 2022, 14, 3425–3440.
44. S. Li, D. Lei, W. Ren, X. Guo, S. Wu, Y. Zhu, A. L. Rogach, M. Chhowalla and A. K.-Y. Jen, Nat. Commun., 2020, 11, 1192.
45. D. He, X. Xu, Z. Liang, Y. Niu, Y. Sun, T. Gavin, P. Falaras and L. Hu, J. Mater. Chem. C, 2021, 9, 9584–9591.
46. V. Gurin, J. Cryst. Growth, 1998, 191, 161–165.
47. A. Shestha, Y. Yin, G. G. Andersson, N. A. Spooner, S. Qiao and S. Dai, Small, 2016, 13, 1602956.
48. M. Zhao, W. Xie and C. Cui, Chem. – Eur. J., 2014, 20, 9259–9262.
49. N. A. DeLucia, N. Das and A. K. Vannucci, Org. Biomol. Chem., 2018, 16, 3415–3418.
50. P. Asgari, Y. Hua, A. Bokka, C. Thiamsiri, W. Prasitwatcharakorn, A. Karedath, X. Chen, S. Sardar, K. Yum and G. Leem, Nat. Catal., 2019, 2, 164–173.
51. A. J. Prochazkova, M. C. Scharber, C. Yumusak, J. Jančík, J. Másilko, O. Brüggemann, M. Weiter, N. S. Sariciftci, J. Krajcovic and Y. Salinas, Sci. Rep., 2020, 10, 15720.
52. A. Mehta, J. Im, B. H. Kim, H. Min, R. Nie and S. I. Seok, ACS Nano, 2018, 12, 12129–12139.
53. A. Liu, C. Bi, J. Li, M. Zhang, C. Cheng, D. Binks and J. Tian, Nano Lett., 2023, 23, 2405–2411.
54. C. Sun, Y. Zhang, C. Ruan, C. Yin, X. Wang, Y. Wang and W. W. Yu, Adv. Mater., 2016, 28, 10088–10094.
55. K. Xu, E. T. Vickers, L. Rao, S. A. Lindley, A. L. C. Allen, B. Luo, X. Li and J. Z. Zhang, Chem. – Eur. J., 2019, 25, 5014–5021.
56. R. M. Pasternack, S. R. Amy and Y. J. Chabal, Langmuir, 2008, 24, 12963–12971.

---

Footnotes

† Electronic supplementary information (ESI) available: Fig. S1–S12 and Movies S1–S3. See DOI: https://doi.org/10.1039/d4nr04890c

‡ These authors contributed equally to this work.

---