Journal of Materials Science & Technology  2020 , 43 (0): 84-91 https://doi.org/10.1016/j.jmst.2019.10.016

Research Article

Effects of 1,9-dibromnonane on the structural, photophysical properties and stability of cesium lead bromide perovskite nanocrystals

Zhaojun Moa*, Qiujie Lua, Zhihong Haob, Zhexuan Zhenga, Fu Qiua, Xiao Yanga, Zhenyu Lia, Lan Lia

a School of Material Science and Engineering, Institute of Material Physics, Key Laboratory of Display Materials and Photoelectric Devices of Ministry of Education of Ministry of Education, Key Laboratory for Optoelectronic Materials and Devices of Tianjin, Tianjin University of Technology, Tianjin 300191, China
b Tianjin Vocational Institute, Tianjin 300350, China

Corresponding authors:   ∗Corresponding author. E-mail address: mzjmzj163@163.com (Z. Mo).

Received: 2019-07-9

Revised:  2019-09-13

Accepted:  2019-10-6

Online:  2020-04-15

Copyright:  2020 Editorial board of Journal of Materials Science & Technology Copyright reserved, Editorial board of Journal of Materials Science & Technology

More

Abstract

The CsPbBr3@Cs4PbBr6 nanocrystals (NCs) could be synthesized by multiple Cs-oleate injections as adding the 1,9-dibromnonane in the reaction solution. The 1,9-dibromnonane could provide Br- ions and the rich Br- ions effectively restrain the generation of CsBr. The Cs4PbBr6 wrapped around the CsPbBr3 NCs and the size of crystalline grain was increased with increasing the Cs-oleate as excess oleylamine. The quantum yield for 14S4DN reached to 99.3 % due to the decrease of defects and the surface passivation of Cs4PbBr6. There are more oleylammonium bromide on the surface of CsPbBr3 NCs as synthesized with 1,9-dibromnonane. The ligand shell and the surface passivation of Cs4PbBr6 restrained the decomposition of surface, consequently improved the stability of moisture and light for CsPbBr3 NCs. When the CsPbBr3@Cs4PbBr6 NCs were immersed in water under UV light (365 nm) for 2 h, the PL intensity could retain 90.4 %, while the 11S1 (traditional CsPbBr3 NCs) was only 10.9 %. It indicated the stability of moisture and light for CsPbBr3 NCs were greatly improved, because Cs4PbBr6 NCs effectively passivated the surface of CsPbBr3 NCs and restrained the generation of traps states.

Keywords: Perovskite ; Nanocrystals ; Luminescence ; Stability

0

PDF (2558KB) Metadata Metrics Related articles

Cite this article Export EndNote Ris Bibtex

Zhaojun Mo, Qiujie Lu, Zhihong Hao, Zhexuan Zheng, Fu Qiu, Xiao Yang, Zhenyu Li, Lan Li. Effects of 1,9-dibromnonane on the structural, photophysical properties and stability of cesium lead bromide perovskite nanocrystals[J]. Journal of Materials Science & Technology, 2020, 43(0): 84-91 https://doi.org/10.1016/j.jmst.2019.10.016

1. Introduction

The lead halide perovskite as multifunctional materials have been attracted much attention due to outstanding optoelectronic characteristics. Inorganic metal halide perovskite (CsPbX3, X= Cl, Br or I) has been extensively applied into a new generation optoelectric devices, such as light emitting diodes (LEDs), lasers, photodetectors and solar cells [[1], [2], [3], [4], [5], [6], [7], [8], [9], [10]]. The CsPbX3 nanocrystals (NCs) were prepared by a hot-injection method with oleylamine (OAM) and oleic acid (OA) as organic ligands [11]. In terms of morphologies, CsPbX3 NCs have been synthesized as cubes, nanoplatelets, nanowires and nanocages [[11], [12], [13], [14], [15]]. For composition adjustment, the study focused on ion exchange reactions, some isovalents and aliovalent cations [[16], [17], [18]]. As a result, the properties of CsPbX3 NCs are strongly dependent on composition and preparation conditions.

According to the connectivity of corner-sharing PbX64- octahedra, perovskites include four crystal structures: 0-dimensional (0-D), 1-dimensional (1-D), 2-dimensional (2-D) and 3-dimensional (3-D) lattices. The low-dimensional-networked perovskites exhibit more stable property [19]. Recently, Cs4PbBr6, a representative 0-D perovskite has attracted much attention, because 0-D perovskite exhibits superior stability and a high photoluminescence quantum yield (PLQY) in the form of both single crystal and powders [[20], [21], [22], [23]]. The solvent in colloidal or the thin film of Cs4PbBr6 NCs perovskite is not sensitive to air environment [21]. The Cs4PbBr6 NCs were prepared by using an excess of Cs-oleate at low reaction temperature [24]. The chemical transformation between Cs4PbBr6 and CsPbBr3 has also been explored. Based on the ligand-assisted supersaturated and recrystallization synthesizing procedure, phase transformation between CsPbBr3 and Cs4PbBr6 was controlled by tuning the amounts of surfactants [25]. Alivisatos group reported a ligand mediated transformation of presynthesized CsPbBr3 NCs to Cs4PbBr6. The transformation was initiated by amine addition, and the use of alkylthiol ligands greatly improved the size uniformity and chemical stability of the derived NCs [26]. Manna group converted Cs4PbBr6 NCs to CsPbBr3 NCs by treating presynthesized Cs4PbBr6 NCs with excess PbBr2 [27,28]. Cs4PbBr6 as a CsBr-rich structure, Yin group synthesized monodisperse CsPbBr3 NCs by a CsBr-stripping process from Cs4PbBr6 NCs precursors [29]. The CsPbBr3 NCs were embedded in Cs4PbBr6 microcrystals by a solution-phase synthesis with high PLQY (90 %) and stability [23]. In previous reports, Cs4PbBr6 NCs were regarded as PbBr2-deficient or CsBr-rich material. Monodisperse CsPbBr3 NCs were synthesized by PbBr2-adding or CsBr-stripping in the Cs4PbBr6 NCs. The CsPbBr3 NCs could also transform to Cs4PbBr6 NCs by adding the Cs- in Br- rich reaction solution.

In this work, we report that the synthesis of a new class of colloidal CsPbBr3@Cs4PbBr6 NCs with high QY. The stability of moisture and light for CsPbBr3 NCs were improved by the decrease of defects and the surface passivation of Cs4PbBr6 as adding 1,9-dibromnonane (1,9-DN) in reaction solution.

2. Experimental

2.1. Materials

Cesium carbonate (Cs2CO3, AR, 99 %, Aladdin), lead bromide (PbBr2 AR, Aladdin), octadecene (ODE, 90 %, Aladdin), oleic acid (OA, AR, Aladdin), oleylamine (OAM, 80-90 %, Aladdin), 1,9-dibromnonane (C9H18Br2 97 %, Aladdin) and toluene (AR, Aladdin) were used without purification unless otherwise noted.

2.2. Preparation of Cs-Oleate

2.5 mmol Cs2CO3, 2.5 mL OA, and 10 mL ODE were loaded into a 50 mL three-neck flask, the reaction solution was stirred at 120 °C for 1 h, then heated under Ar to 150 °C until all Cs2CO3 reacted with oleic acid.

2.3. Preparation of PbBr2 solution

ODE (25 mL) and PbBr2 (0.8 mmol) were added into a 100 mL three-neck flask, the reaction solution was stired at 120 °C for 1 h, then added OAM and OA under Ar2. Next, temperature was raised to 180 °C for 30 min, and the PbBr2 salt completely dissolved. The 1,9-DN was added and keep several minutes or without 1,9-DN.

2.4. Synthesis of CsPbBr3 NCs

Firstly, the Cs-oleate solution (1 mL, 0.4 M) was quickly injected. The reaction mixture was kept at 180 °C for 5 min and then extracted out a quarter of solution as S1. Next, the Cs-oleate solution (0.75 mL, 0.4 M) was quickly injected and kept at for 5 min, then extracted out one third of solution as S2. Thirdly, the Cs-oleate solution (0.5 mL, 0.4 M) was quickly injected and kept at for 5 min, then extracted out a half of solution as S3. Finally, the Cs-oleate solution (0.25 mL, 0.4 M) was quickly injected and kept at for 5 min as S4. All the samples were cooled to room temperature by an ice-water bath.

The final NCs were purified by centrifugation and redispersed in toluene to form a stable solution. The samples were named according to the preparation conditions as shown in Table S1.

2.5. Characterizations

The structure of samples was determined by X-ray diffraction (XRD) (Rigaku D/max 2500 v/pc diffractometer) with CuKα radiation. The morphologies were recorded by a field emission scanning electron microscopy (SEM, HITACHI S-8040) and transmission electron microscopy (TEM, JEM-2100 F). The absorption spectra were characterized on a Hitachi UV-4100 UV-vis spectrophotometer. Photoluminescence (PL) emission spectra and QY were measured by Horiba JYFL3 fluorescence spectrometer. X-ray photoelectron spectroscopy (XPS) (VGESCALAB MKII instrument with Mg Ka ADES source) was used to perform elemental analysis.

3. Results and discussion

3.1. Morphology and structure

Fig. 1 shows the XRD patterns of the samples synthesized without 1,9-DN. The peaks at 2θ = 15.16°, 21.52°, 30.63°, 37.58°, and 43.52° can correspond to the diffractions from (100), (110), (200), (210), (211), and (202) planes of crystalline cubic CsPbBr3 for S1 as shown in the Fig. 1(a)-(c). The diffraction peaks of Cs4PbBr6 were observed as more OA than OAM in solution, and full width at the half maximum (FWHM) of 41S1 is narrower than others as shown in Fig. 1(d). It means the crystalline grain is larger as rich OA in the solution. Cs4PbBr6 were observed in the S2 and the diffraction peaks of Cs4PbBr6 for 14S2 and 41S2 were stronger than 12S2 and 11S2, which indicated the Cs4PbBr6 phase could easily generate as rich OA or OAM in the solution. With increasing the Cs-oleate, the composition of samples were different as different proportion of OA:OAM. Fig. 1(a) and (b) shows the strong diffraction peaks of CsBr for 14S3 and 12S3. And, the diffraction peaks of CsBr were increased and CsPbBr3 were decreased for 14S4 and 12S4, with increasing the content of Cs-oleate. When OA was equal or rich to OAM in solution, the diffraction peaks of CsBr almost were not observed for 11S3 and 41S3 as shown in Fig. 1(c) and (d). For 11S4, the diffraction peaks of Cs4PbBr6 could still be observed, but the CsPbBr3 was not. Although the CsBr was the main phase, Cs4PbBr6 and CsPbBr3 were all observed in 41S4. It indicated the CsPbBr3 can continue to react with Cs-oleate and synthesize CsBr and the ratio of OA:OAM is very important for reaction process. The CsPbBr3 was probably degraded due to a reaction with the excess of amine [30]. Therefore, the synthesis process could be described by the reaction equation:

PbBr2 + 2HOOCR1 + 2R2NH2 → Pb(OOCR1)2 + 2R2NH3Br (HOOCR1 = oleic acid R2NH2 = oleylamine) (1)

CsOOCR1 + Pb(OOCR1)2 + 3R2NH3Br → CsPbBr3↓+ 3R2NH3OOCR1 (2)

The CsPbBr3 NCs were synthesized with injecting the Cs-oleate solution: 3CsOOCR1 + CsPbBr3↓+ 3R2NH3Br→ Cs4PbBr6↓ + 3R2NH3OOCR1 (3)

Fig. 1.   XRD patterns of the CsPbBr3 nanocrystals synthesized without 1,9-DN under different ratios of OA and OAM: (a) OA:OAM = 1:4; (b) OA:OAM = 1:2; (c) OA:OAM = 1:1; (d) OA:OAM = 4:1.

The Cs4PbBr6 NCs was synthesized with injecting the Cs-oleate solution.

2CsOOCR1 + CsPbBr3↓→3CsBr↓+ Pb(OOCR1)2 (4)

The CsBr was synthesized with injecting the Cs-oleate solution.

The majority of CsPbBr3 NCs for 14S1, 12S1 and 11S1 exhibited cube with 10-15 nm as shown in Fig. S1(a)-(c) in Supporting information. The bigger cubes were observed in 14S2, 12S2 and 11S2, indicating the CsPbBr3 NCs could grow up with increasing Cs-oleate. And then the NCs were disappeared in the 14S3, but a few cubic NCs still were observed in the 12S3 and 11S3, they evolved into nonuinform cubes with increasing the Cs-oleate. Fig. S1(d) shows absolutely different evolution process as the excess OA in the reaction solution. The morphology of 41S1 is nonuinform cubes due to the coexistence of Cs4PbBr6 and CsPbBr3, the majority of Cs4PbBr6 NCs for 41S2 exhibited 50 nm, and the morphology of 41S3 is nonuinform again with appearing CsBr. Finally, the majority of cubic CsBr for 41S4 is 200-500 nm. Corresponding to the XRD, the evolution process of the morphology is closely related to the reaction process.

When the 1,9-DN was added in the reaction solution, the development process was very different. Although the XRD patterns of the S1-DN exhibited cubic CsPbBr3, the FWHMs were different. When OA was equal or less to OAM, the FWHM was wider than rich OA as shown in Fig. 2(a)-(d). It indicated the excess of OA could promote the growth of CsPbBr3 NCs. What's more, the diffraction peaks of Cs4PbBr6 were only observed in 14S2-DN and 41S2-DN in Fig. 2(a) and (d). It implied the excess of OA or OAM could accelerate the reaction to generate Cs4PbBr6. Compared to 12S2 and 11S2, the 1,9-DN could restrain the generation of Cs4PbBr6. Continuing to increase the content of Cs-oleate, the diffraction peaks of CsBr were not observed for S3-DN and S4-DN, suggesting the CsBr could be effectively restrained by rich Br- coming from 1,9-DN. The process could be described by the reaction equation:

BrR3Br + 4R2NH2 → R2NHR3NHR2 + 2R2NH3Br (5)

(BrR3Br = 1,9-DN)

When the OAM was excess, the FWHM was broadened with increasing the content of Cs-oleate from S1-DN to S4-DN as shown in Fig. 2(a) and (d). Compared to the S2 synthesized without 1,9-DN (Fig. 1(a) and (b)), the evolvement process was completely different. It implied that the growth of CsPbBr3 NCs could be efficiently inhibited as adding 1,9-DN in the reaction solution. Whereas, when the OA was equal or more to OAM, the FWHM was narrowed with increasing the content of Cs-oleate from S1-DN to S4-DN as shown in Fig. 2(c) and (d). It indicated the OA could promote to grow up for CsPbBr3 NCs.

Fig. 2.   XRD patterns of the CsPbBr3 nanocrystals synthesized with 1,9-DN under different ratios of OA and OAM: (a) OA:OAM = 1:4; (b) OA:OAM = 1:2; (c) OA:OAM = 1:1; (d) OA:OAM = 4:1.

The evolution process of morphology was also very different, when the 1,9-DN was added in the reaction solution. When the OAM is excess, the morphology of samples almost was unchanged from 14S1-DN and 12S1-DN to 14S2-DN and 12S2-DN as shown in the Fig. S2(a) and (b). It was different with the 12S2 and 14S2, implying that the growth of CsPbBr3 NCs could be efficiently inhibited as adding 1,9-DN in the reaction solution. With increasing the content of Cs-oleate, the Cs4PbBr6 was generated. Some larger nanoparticles (NPs, ∼20 nm) with a spherical shape are observed in the 14S3-DN, 12S3-DN, 14S4-DN and 12S4-DN as shown in Fig. S2(a) and (b). The size of cubic CsPbBr3 NCs for 14S1-DN exhibited cubic with an average side length of ∼10 nm as shown in Fig. S3(c). When the Cs-oleate was added, the morphology was not changed, but the size of CsPbBr3 NCs (14S2-DN) was larger as shown in Fig. S3(d). Fig. S3(e) shows some larger nanoparticles (NPs, 15-20 nm), which was determined to CsPbBr3/Cs4PbBr6. Continue increasing the content of Cs-oleate, the proportion of spherical shape was increased as shown in Fig. S3(f). The insets showed the high-resolution TEM (HRTEM) images. The CsPbBr3@Cs4PbBr6 looked like the core-shell structure, Cs4PbBr6 wrapped around the CsPbBr3 NCs. Although the size of samples was increased, the FWHM of XRD was broadened with increasing the content of Cs-oleate for 14S3-DN and 14S4-DN as shown in Fig. 2. It indicated that Cs4PbBr6 layer was formed by consuming CsPbBr3 NCs and this layered structure was formed from the outside to the inside, which leaded to decreasing the size of CsPbBr3. The CsPbBr3 NCs of 11S1 and 11S1-DN exhibited cubic with an average side length of ∼10 nm as shown in the Fig. S3(a) and (b), which was similar to 14S1-DN. However, when the OA was equal to OAM, the size of CsPbBr3 NCs was increased and appeared the big cubes with increasing the content of Cs-oleate from 11S1-DN to 11S4-DN as shown in Fig. S2(c). Although the XRD indicated the 41S1-DN was pure phase of CsPbBr3, the size of cubic CsPbBr3 was not uniform as shown in the Fig. S2(d). The boundary of the cube gradually blurred and changed into irregular particles with increasing the content of Cs-oleate from 41S1-DN to 41S4-DN. It indicated the excess OAM could improve to synthesize the uniform cubic CsPbBr3 NCs.

3.2. Optical property

Fig. 3(a)-(d) shows the UV-vis absorption spectrum of samples synthesized without 1,9-DN. The absorption spectra of S1 were obviously observed about 510 nm, which belonged to CsPbBr3 NCs as shown in Fig. 3(a)-(c). Although the XRD indicated the 41S1 was almost CsPbBr3 phase and trace of Cs4PbBr6 as excess OA, absorption of 41S1 was observed about 320 nm rather than 510 nm in the UV-vis absorption spectrum as shown in the Fig. 3(d). It indicated the Cs4PbBr6 dispersed in the toluene, which might be attributed to the bigger size of CsPbBr3. The absorption peaks 320 nm for S2 and S3 were observed due to generating the Cs4PbBr6. The profile of the absorption spectra for samples synthesized with 1,9-DN was totally same above 350 nm as shown in Fig. 3(e) and (f). But the absorption peak of 320 nm (3.87 eV) continued to strengthen from S1-DN to S4-DN, which was attributed to the increase of Cs4PbBr6 phase. The insulator band gap of Cs4PbBr6 phase is 3.9 eV [27]. When the OA was equal to OAM, although Cs4PbBr6 phase was generated, the absorption of 320 nm for 11S3-DN was not increased due to the bigger size as shown in the Fig. 3(g). The weak absorption of 41S-DN may be attributed to low concentration due to the bigger size based the SEM as shown in Fig. 3(h).

Fig. 3.   UV-vis absorption spectrum of the CsPbBr3 nanocrystals under different ratios of OA and OAM: (a, e) OA:OAM = 1:4; (b, f) OA:OAM = 1:2; (c, g) OA:OAM = 1:1; (d, h) OA:OAM = 4:1.

Fig. 4(a)-(d) shows the PL spectra of the samples synthesized without 1,9-DN. The PL spectra of S1 showed a narrow emission with FWHM ∼18 nm (∼84 meV) as shown in Fig. 4(a)-(c). The band center is about 519 nm, implying the direct exciton recombination luminescence of the CsPbBr3 NCs [[9], [10], [11], [12]]. The weak PL intensity of 14S2, 12S2, 11S2 and 41S1 might be attributed to the bigger grain based XRD and SEM. CsPbBr3 continued to react with Cs-oleate and synthesized CsBr, so the PL spectrums were not observed in the S3 and S4. When the 1,9-DN was added in the reaction solution, the evolution process of phase and morphology were completely different, therefore, the PL spectrums were also different. Compared to 11S1, the PL intensity of 11S1-DN was improved, when the OA is equal to OAM. It may be caused by decreasing the Br vacancy (VBr) due to additional Br- (adding 1,9-DN). The additional Br- was introduced to drive the ionic equilibrium to form intact Pb-Br octahedrons and reduced nonradiative recombination processes [33]. But, the PL intensity decreased from 11S1-DN to 11S4-DN due to the growth of CsPbBr3 NCs with increasing the content of Cs-OA as shown in Fig. 4(g). Because the 1,9-DN would consume much OAM and lead to the imbalance of OA and OAM in the solution, which resulted in the growth of CsPbBr3 NCs. The PL intensity of samples could be improved as excess OAM in the reaction solution as shown in Fig. 4(e) and (f). What’s more, the PL intensity was not decreased with increasing the content of Cs-OA (from S1-DN to S4-DN) as excess OAM, because the Cs4PbBr6 wrapped around the CsPbBr3 NCs, and effectively prevented the growth of CsPbBr3 NCs. The Cs4PbBr6 layers were formed from the outside to the inside [25,30]. The PL intensity was very weak as excess OA in the reaction solution as shown in Fig. 4(h). It may be attributed to the larger grain.

Fig. 4.   PL spectrums of the CsPbBr3 nanocrystals synthesized without and with 1,9-DN under different ratios of OA and OAM: (a, e) OA:OAM = 1:4; (b, f) OA:OAM = 1:2; (c, g) OA:OAM = 1:1; (d, h) OA:OAM = 4:1.

In order to analysis the carrier recombination dynamics of the CsPbBr3 NCs, the time resolved PL decay was measured. Fig. S4 shows the PL decay curves for the CsPbBr3 NCs, which could be well fitted with a Bi-exponential function:

A(t)=A1exp($-\frac{t}{τ_{1}}$)+A2exp(-$\frac{t}{τ_{2}}$) A(t)=A1exp(-$\frac{t}{τ——{1}}$ )+A2exp(-$\frac{t}{τ_{2}}$) (6)

where A, A1 and A2 are constants, t is time, and τ1 and τ2, represent the decay lifetimes corresponding the recombination of initially generated excitons upon light absorption and the excitonic recombination with the involvement of shallow levels defects and surface states. And the average lifetimes were calculated using the following equation:

τave=(A1$τ_{1}^{2} + A_{2}τ_{2}^{2}$)/(A1τ1+A2τ2) (7)

The short-lived PL lifetime (τ1) about 6 ns could correspond to the recombination of excitons and a long lived PL lifetime (τ2) about 24 ns may be ascribed to the excitonic recombination with the involvement of shallow levels defects and surface states as shown in Table S2. The percentage of τ2 (30.38 %) for 11S1 decreased to 23.25 % for 11S1-DN as adding 1,9-DN, indicating the reduce of the shallow levels defects (VBr). The percentage of τ2 for 14S1-DN is 36.38 %, which may be attributed to the rich surface states as excess OAM. Because the Oleylammonium bromide and protonated OAM could draw OA into the ligand shell, the dynamic surface was wrapped by oleylammonium bromide, OAM and OA. The percentage of τ2 was reduced due to the surface passivation (Cs4PbBr6 wraps around the CsPbBr3 NCs) with increasing the content of Cs-oleate (from 14S1-DN to 14S4-DN).

The absolute PLQY of the CsPbBr3 NCs solution was determined by using a fluorescence spectrometer with an integrated sphere with the excitation wavelength of 397 nm. Fig. S5 shows the PL of the CsPbBr3 NCs solution, which were used to calculate the absolute PLQY. The PLQY of S1 was decreased from 51.07 % to 39.07 % with increasing the OAM (from 11S1 to 14S1) under without 1,9-DN in the reaction solution as shown in Fig. 5. The CsPbBr3 NCs might be degraded and increased the defects as the excess of amine [31]. Although the CsPbBr3 NCs had high defect tolerance, the increase of VBr density would increase nonradiative recombination processes [32,33]. When the OA is equal to OAM, the PLQY was improved from 51.07 % (11S1) to 72.48 % (11S1-DN) as adding the 1,9-DN. The increase may be attributed to the decrease of VBr, because the 1,9-DN could provide rich Br-. The PLQY was further improved from 72.48 (11S1-DN) to 90.89 % (14S1-DN) with increasing the OAM. The 1,9-DN couldn't directly provide the activation of bromide ions, it should be activated by OAM using the reaction (5). Therefore, the 1,9-DN could provide more Br- to reduce the defect of CsPbBr3 NCs, and the balance between OA and OAM could guaranteed to restrain the growth as increasing the OAM. Additionally, when the OAM is more than OA, the PLQY was improved with increasing the content of Cs-oleate, and the PLQY of S4-DN reached to 95.33 % and 99.28 %. The increase of PLQY may be attributed to the decrease of defects and the surface passivation of Cs4PbBr6 [23,25,26,30].

Fig. 5.   Absolute PLQY of the CsPbBr3 NCs solution.

In order to further confirm the change of defects and the surface passivation, the XPS spectra were measured as shown in Fig. S6 and the binding energies and areas of Pb 4f exhibited in Table S3. Two peaks of Pb 4f5/2 and Pb 4f7/2 in XPS core level spectra are defined [37]. The peaks of Pb 4f5/2 and Pb 4f7/2 for 11S1 were fitted with two peaks. The high binding energies (142.99 and 138.10 eV) were assigned to Pb-Br, and the low binding energies (142.62 and 137.73 eV) were assigned to Pb-oleate. The Pb-oleate signal indicates the existence of VBr defect [38]. But the Pb 4f5/2 and Pb 4f7/2 of 11S1-DN could hardly be fitted into two peaks, which blue shift to higher binding energies of 143.00 and 138.12 eV, corresponding with the Pb-Br bonding well. It indicated that the VBr was succcessfully eliminated by adding 1,9-DN. When the OAM is more than OA (4:1), the low binding energies (142.62 and 137.73 eV) were observed again, which maybe cause by more protonated OAM draws OA into the ligand shell. Finally, the low binding energies disappeared in 14S4-DN due to passivation of Cs4PbBr6 with increasing Cs-oleate. Trap (VBr) density could calculated by peak areas of Pb-Br and Pb-oleate using the relation:

$\frac{ n_{1}}{ n_{2}}=\frac{ I_{1}/S_{1}}{ I_{2}/S_{2}}$ (8)

where the n is mole, the I is peak area, the S is sensitivity factor. The peak areas were listed in the table. S3. The Pb-Br/Pb-oleate of 11S1 was calculated to be 3.53/1, which implied the trap density was 22.1 %. And, the trap densities were 6.7 %, 6.6 % and 5.2 % for 14S1-DN, 14S2-DN and 14S3-DN, respectively.

3.3. Stability

Fig. 6(a1) shows the PL intensity of 11S1 reduced to 50.4 % under UV light (365 nm) for 2 h. The PL spectrums of CsPbBr3 NCs dispersed in to poly(methyl methacrylate) (PMMA) and immersed in water were measured as shown in Fig. 6(a2). The PL intensity of 11S1 reduced to 25.7 % after 2 h. And when the 11S1 were immersed in water under UV light (365 nm), the PL intensity decreased to 10.9 % after 2 h as shown in Fig. 6(a3). It indicated the PL intensity was accelerated to decrease as immersing in water under UV light. The 11S1 exhibited a weak stability, because the surface ligands (OA and OAM) are not tightly bound to the surface of CsPbBr3 NCs, they would be easily removed and the surface of the CsPbBr3 NCs could be decomposed, which led to increasing trap states [34]. The PL intensity of 11S1-DN reduced to 61.5 % under UV light (365 nm) for 2 h as shown in Fig. 6(b1). When the CsPbBr3 NCs immersed in water after 2 h, the PL intensity was 52.4 % as shown in Fig. 6(b2). And when the CsPbBr3 NCs were immersed in water under UV light (365 nm) for 2 h, the PL intensity decreased to 36.5 % as shown in Fig. 6(b3). Compared to 11S1, the stability of the 11S1-DN was enhanced. When the CsPbBr3 NCs were synthesized with 1,9-DN, the dynamic surface was stabilized by oleylammonium bromide, which came from Eq. (5). While protonated OAM drew OA into the ligand shell, which could slow the decomposition of surface and the generation of traps states, so the dynamic surface was stabilized by oleylammonium bromide, OA and OAM [35]. But, the stability was also weakest as immersing in water under UV light. It implied the synergistic effect among illumination and moisture could accelerate the degradation of CsPbBr3 NCs. Fig. S7(a) and (c) exhibites the cubic CsPbBr3 NCs (∼14 nm) without exposed UV light (365 nm). The size of CsPbBr3 NCs is obviously decreased, and the edges of cubic CsPbBr3 NCs was dim under UV light (365 nm) for 2 h as shown in Fig. S7(b) and (d). It implies the surface of CsPbBr3 NCs was seriously decomposed. The decomposition of surface and the increase of traps states play an important role for PL loss [36]. Therefore, the decrease of PL intensity mainly came from the decomposition of surface. Although, the stability of the CsPbBr3 NCs was improved by adding 1,9-DN, it did still not effectively prevent the decomposition of surface.

Fig. 6.   PL intensity of CsPbBr3 NCs were exposed UV light (365 nm), immersed in water and synergistic effect for 2 h: (a1-a3): 11S1; (b1-b3): 11S1-DN.

When the OAM was more than OA (OA:OAM = 1:4), the stability of CsPbBr3 NCs was greatly improved. Fig. 7(a1)-(d1) shows the PL spectrs of CsPbBr3 NCs under UV light (365 nm) for 2 h. The PL intensity remained 83.2 %, 91.0 %, 90.6 % and 98.5 % for the CsPbBr3 NCs (from 14S1-DN to 14S4-DN). Compared to the CsPbBr3 NCs (11S1) synthesized without 1,9-DN 50.4 %, the stability was greatly enhanced. The enhancement of stability may be attributed to the surface passivation of CsPbBr3 NCs. The PL spectra of CsPbBr3 NCs dispersed in to poly(methyl methacrylate) (PMMA) and immersed in water were measured. Fig. 7(a2)-(d2) shows the PL spectra of CsPbBr3 NCs immersed in water for 2 h. The PL intensity reduced to 72.5 %, 73.8 % and 76.5 % for the 14S1-DN, 14S2-DN and 14S3-DN. The stability of moisture for the CsPbBr3 NCs was greatly enhanced. Especially, When the CsPbBr3 NCs was completely wrapped by Cs4PbBr6, the PL intensity remained 97.5 % for 14S4-DN. When the CsPbBr3 NCs were immersed in water under UV light (365 nm) for 2 h, the PL intensity of the CsPbBr3 NCs reduced to 52.3 % 68.7 % and 71.5 %, and the PL intensity is only reduced 9.6 % for14S4-DN as shown in Fig. 7(a3)-(d3). Compared to 11S1-DN, the illumination and moisture stability of 14S-DN was improved. It maybe cause by more protonated OAM ligands on the surface of CsPbBr3 NCs such as ligand shell, which can prevent the decomposition of surface and restrain the generation of traps states. And the illumination and moisture stability were further improved, when the Cs4PbBr6 was coated on the surface of CsPbBr3 NCs. The Cs4PbBr6 exhibited superior stability, which could effectively passivate the surface of CsPbBr3 NCs and restrain the decomposition of surface as well as the increase of traps states. Fig. S8 exhibited the cubic CsPbBr3 NCs exposed UV light (365 nm) for 2 h. The size of 14S1-DN was little decreased as shown in Fig. S8(a), but the size of 14S2-DN, 14S3-DN and 14S4-DN was unchanged as shown in Fig. S8(b)-(d). It implied the surface of CsPbBr3 NCs was effectively protected by Cs4PbBr6.

Fig. 7.   PL intensity of CsPbBr3 NCs immersed in water, exposed UV light (365 nm) and synergistic effect for 2 h: (a1-a3) 14S1-DN; (b1-b3) 14S2-DN; (c1-c3) 14S3-DN; (d1-d3) 14S4-DN.

4. Conclusion

The CsPbBr3 NCs could be transformed into Cs4PbBr6 by multiple Cs-oleate injections, and the content of OA, OAM and Br- played important roles. The excess OA or OAM could promote to generate Cs4PbBr6, and the rich Br- ions could effectively restrain the generation of CsBr. With increasing the Cs-oleate, the size of CsPbBr3 NCs was increased, but the 1,9-DN could effectively restrain the growth of CsPbBr3 NCs as excess OAM. What’s more, when the OAM is more than OA, the PLQY was greatly improved due to the decrease of defects as adding the 1,9-DN in the reaction solution. And the PLQY could be further increased by adding the Cs-oleate, which may be attributed to the decrease of defects and the surface passivation of Cs4PbBr6. The XPS confirmed the decrease of defects by adding the 1,9-DN. Meanwhile, the Cs4PbBr6 wrapped around the CsPbBr3 NCs. This structure greatly improved the stability of moisture and light for CsPbBr3 NCs. For the conventional CsPbBr3 NCs (11S1), when they were immersed in water under UV light (365 nm) for 2 h, the PL intensity was only 10.9 %, but the PL intensity of CsPbBr3@Cs4PbBr6 (14S4-D N) could remain 90.4 %. The improvement of the moisture and light stability for CsPbBr3 NCs was attributed to the surface modification. There are more oleylammonium bromide on the surface of CsPbBr3 NCs as synthesized with 1,9-DN, the protonated OAM draws OA into the ligand shell, and restrained the decomposition of surface and the generation of traps states. And, Cs4PbBr6 could more effectively passivate the surface of CsPbBr3 NCs.

Acknowledgments

This work was supported financially by the National Natural Science Foundation of China (Nos. 11504266, 51702235 and 51871167), the Tianjin Natural Science Foundation (No. 17JCQNJC02300), the Science & Technology Development Fund of Tianjin Education Commission for Higher Education (No. 2017KJ247) and the National Key Foundation for Exploring Scientific Instrument of China (No. 2014YQ120351).


/