Synthesis of all-inorganic CsPb₂Br₅ perovskite and determination of its luminescence mechanism

Abstract

All-inorganic cesium lead halide perovskite materials, such as CsPbBr₃ and CsPb₂Br₅, have drawn immense attention recently due to their potential applications in optoelectronics. However, unlike the extensively-studied CsPbBr₃, many of the fundamental properties of CsPb₂Br₅ remain unclear, such as its electronic structure and photoluminescence (PL) mechanism. In this work, we developed a novel environmentally friendly chemical method to synthesize CsPb₂Br₅, where only water and ethanol were used as solvents. The optical property and electronic structure of CsPb₂Br₅ were then examined both experimentally and theoretically. For the first time we found that the band-edge emission of CsPb₂Br₅ was not in the visible range as reported in the literature but rather located in the ultraviolet range. By monitoring the PL property during phase conversion between CsPb₂Br₅ and CsPbBr₃, we propose that previously-reported visible emission might be due to the involvement of CsPbBr₃ during sample preparation. Overall, this study presents a novel green method for the preparation of CsPb₂Br₅ and clarifies the optical properties and electronic structure of CsPb₂Br₅.

---

1. Introduction

In the recent years, newly emerging metal halide perovskite crystals, in particular organic–inorganic perovskites such as CH₃NH₃PbX₃ (X = Cl, Br, and I), have attracted tremendous scientific attention, thanks to the rapid development of perovskite solar cells whose power conversion efficiency already exceeds 20% since the first report of 3.81% in the year of 2009.^1,2 In addition to the applications in photovoltaics, perovskites are also reported in high-efficiency light emitting diodes (LEDs), lasers and photodetectors. From the standpoint of practical application, however, hybrid organic–inorganic perovskites suffer from low environmental- and photo-stability due to their undesirable reaction with water and degradation under UV light irradiation.^3–5 Perovskite solar cells start to degrade within several hours without encapsulation; the chemical processes that cause the degradation of perovskite films were reviewed by Niu et al.⁶ Many efforts are being attempted to mitigate this degradation issue but with very limited success.⁷ As a type of perovskite material, the all-inorganic perovskites have recently been proposed as promising candidates for optoelectronics because of their higher chemical stability as compared to their hybrid organic–inorganic counterparts.^8–10 Cesium lead halide (CsPbX₃, X = Cl, Br, and I) is the most widely reported material, which was typically prepared in the range of 140–200 °C by wet chemical method.¹¹ A variety of optoelectronic devices based on CsPbX₃ have been reported.

Very recently, a new type of inorganic perovskite material, CsPb₂Br₅, has been found to also exhibit superior optical property and potential application in optoelectronic device.^12–15 Especially interesting is that unlike CsPbBr₃ that requires high temperature for the synthesis, CsPb₂Br₅ can be prepared in a facile way at room temperature, which is very attractive for the future applications. Yu et al. first reported on the synthesis of CsPb₂Br₅ and its efficient photoluminescence (PL) in the visible region (512 nm) with a quantum yield of 87%.¹² Zhou et al. demonstrated visible amplified spontaneous emission (ASE) from CsPb₂Br₅ microplates with very low pump threshold.¹³ Sun et al. reported the formation of CsPb₂Br₅ as a by-product during the preparation of CsPbBr₃ and also noted an enhancement in the light emission efficiency of CsPbBr₃ LED due to the generation of CsPb₂Br₅.¹⁴ However, there is a debate about the luminescence mechanism of inorganic CsPb₂Br₅ perovskite. Based on the density functional theory (DFT) simulation as well as experimental work, it has been noted that the visible light emission gradually decreases and ultimately disappears accompanying the phase conversion from CsPbBr₃ to CsPb₂Br₅. Jiang et al. emphasized that CsPb₂Br₅ is an indirect bandgap semiconductor and PL-inactive.¹⁵ Therefore, fundamental research on the optical property and electronic structure of CsPb₂Br₅ is imperative, which is important to explain the current controversial experimental observations and to explore its mechanism of luminescence.

In this context, we report a facile method on the fabrication of tetragonal perovskite, CsPb₂Br₅ at ambient temperature as well as its phase conversion between CsPb₂Br₅ and CsPbBr₃. By precisely controlling the phase conversion and monitoring the evolution of PL property, we find that CsPb₂Br₅ has a bandgap of 3.13 eV and correspondingly exhibits a band-edge emission in the ultraviolet region, and hence, the frequently observed visible PL peak can be attributed to the CsPbBr₃ by-product in CsPb₂Br₅. These results not only clarify the luminescence mechanism of CsPb₂Br₅ being under debate, but also provide valuable insights for the development of CsPb₂Br₅ based photoelectric devices.

2. Experimental

Synthesis of CsPb₂Br₅

CsPb₂Br₅ nanoplates were synthesized by chemical precipitation conducted at room temperature. Firstly, purified water and ethanol were mixed to obtain the solvent mixture (total volume = 2 ml, V_H2O : V_EtOH = 1 : 3). CsAc (1 mmol) was added into the above solution and dissolved completely. PbBr₂ (2.5 mmol) was added into CsAc solution and mixed mechanically which resulted in the formation of CsPb₂Br₅ as a precipitate at the bottom of the solution. The precipitate was further treated with HBr to remove the impurities and then washed with water.

Characterization of CsPb₂Br₅

The crystal structure was measured by using a X-ray diffractometer (Bruker D8 Advance) with Cu-Kα radiation (λ = 1.5406 Å). The SEM and EDS were measured using JSM7100F. Absorption spectra were obtained by UV-Visible-NIR Spectrophotometer (SHIMADZU UV-3600). Edinburgh FLS 980 was used to measure the steady-state fluorescence spectrum.

First principle calculations

First-principle calculations were performed on the basis of density functional theory (DFT) as implemented in the QUANTUM ESPRESSO code. The exchange and correlation terms were described using general gradient approximation (GGA) of Perdew–Burke–Ernzerhof (PBE). The energy cutoff for the plane wave basis set was 600 eV. The accuracy of self-consistent field (SCF) energy convergence and the convergence accuracy of the internal stress of crystal is less than 1.4 × 10⁻⁵ eV per atom and 0.05 GPa, respectively. For the different alloy configurations, Monkhorst–Pack grids were determined automatically for Brillouin zone integration and KPPRA parameter was set to 1000.

3. Results and discussion

The CsPb₂Br₅ was synthesized via a chemical precipitation method and the details are described in the Experimental Section. Briefly, cesium acetate (CsAc) was first dissolved in a mixed solvent of purified water and ethanol and PbBr₂ solid was added into the above solvent mixture. As shown by the video (Movie S1) in the ESI,† the reaction is very fast and efficient and results in the formation of CsPb₂Br₅ just several seconds after the addition of PbBr₂ which “emits” green light under UV light irradiation. As CsPb₂Br₅ does not dissolve well in water, a precipitate was obtained at the bottom. The obtained precipitate was further treated with HBr to remove the impurities and then washed with water. We propose the following equations representing the chemical reactions involved in the synthesis, where CsAc is hydrolyzed at first to generate CsOH, which then reacts with PbBr₂:

CsAc + H₂O → CsOH + HAc (1)

CsOH + PbBr₂ → CsPb₂Br₅ + Pb(OH)Br (2)

Pb(OH)Br + HBr → PbBr₂ + H₂O (3)

To investigate the possibility of eqn (2), which is crucial for the formation of CsPb₂Br₅, the total energy (E_t) of the chemical reaction was calculated by employing first principles. The change in total energy (ΔE_t) of −4.558 eV indicates that the reaction is exothermic and should occur spontaneously. The details of calculations can be found in the ESI (Table S1†). The chemical reaction eqn (1) was confirmed experimentally that the PH value was changed to 9 after the CsAc is added to the mixture solvent of water and ethanol.

Fig. 1a shows the scanning electron microscopy (SEM) image, which illustrates that the formed CsPb₂Br₅ nanoparticles are prone to aggregate to form a cluster. Energy dispersive spectroscopy (EDS) analysis (Fig. S1 in the ESI†) reveals the molar ratio of Cs/Pb/Br to be 1/2.1/5.9, which agrees well with the stoichiometry of the compound.

Fig. 1 (a) SEM images of CsPb₂Br₅ cluster. (b) PXRD pattern of tetragonal CsPb₂Br₅ particles (black dots represent PbBr₂). The intensity is in logarithmic scale. (c) Low-resolution, and (d) high resolution lattice resolved TEM image of a representative CsPb₂Br₅ cluster.

The crystalline structure was characterized by powder X-ray diffraction (PXRD) (Fig. 1b). As shown in this figure, all the diffraction peaks match well with the standard tetragonal phase of CsPb₂Br₅ (PDF#25-0211). The presence of small amounts of PbBr₂ has also been noted. The microstructure of CsPb₂Br₅ was further characterized by TEM, as shown in Fig. 1c and d. High resolution TEM reveals the crystalline CsPb₂Br₅ phase with the lattice fringe of 0.744 nm, corresponding to the (002) plane of CsPb₂Br₅. This is well consistent with the XRD results.

Fig. 2a–e illustrate the detailed structure of CsPb₂Br₅. The tetragonal phase of CsPb₂Br₅ belongs to the space group of I4/mcm and it exhibits a sandwich structure consisting of [Pb₂Br₅]⁻ layers and intercalated Cs⁺. The Cs⁺ ions stay in the gaps between the two rectangular pyramid grooves, which have a staggered arrangement. There are two types of bromine atoms occupying different positions in the cell which combine with lead atoms forming a Pb–Br coordination polyhedron. One lead atom and eight bromine atoms make up a hendecahedron. Combining ball-and-stick models and polyhedral models exhibit the tetragonal phase of CsPb₂Br₅ with an interesting sandwich structure. DFT-based first-principles calculations (see details in the Experimental Section) determined the lattice constants as: a = b = 0.86 nm and c = 1.59 nm.

Fig. 2 Crystal structure of tetragonal CsPb₂Br₅. (a), (b) Top and side views of ball-and-stick models for CsPb₂Br₅ unit cell. (c), (d) Top and side views of polyhedral models for the tetragonal structure of CsPb₂Br₅, with sandwiched structure by alternating Cs⁺ and (Pb₂Br₅)⁻ layers. (e) Schematic of Pb–Br hendecahedron (red ball: Cs⁺, green ball: Pb²⁺, blue ball: Br⁻).

The optical absorption and emission spectra of CsPb₂Br₅ are shown in Fig. 3a. An obvious absorption edge at 380 nm could be noted which is different from the previously reported value of 520 nm.^12,13 In the PL spectrum, the two luminescence peaks at 385 nm and 460 nm were detected which exhibit a small Stokes shift as compared to absorption. Fig. 3b gives the plots of (Ahν)^1/2 vs. photo energy, from which a bandgap of 3.13 eV can be found for CsPb₂Br₅.

Fig. 3 (a) Absorption (black) and PL spectra (blue) of CsPb₂Br₅ particles, the PL spectrum under 250 nm excitation. (b) Corresponding (Ahν)^1/2 vs. energy (hν) curve of CsPb₂Br₅. The optical band gap of CsPb₂Br₅ is measured to be ∼3.13 eV.

To reconcile the absorption and emission properties of CsPb₂Br₅ observed here with previous reports (band edge at ∼515 nm), we scrutinized the absorption spectrum and indeed found another weak absorption edge at ∼515 nm as highlighted with the red circle in Fig. 3a. Since two absorption edges were observed, a reasonable assumption is that the one observed at 515 nm is due to an impurity, the most possible origin of which would be CsPbBr₃ considering that there exists a phase transition between CsPbBr₃ and CsPb₂Br₅ as well as that the absorption edge of CsPbBr₃ is well known to be located at 515 nm and PbBr₂ does not show light emission at room temperature.^16,17

To confirm the above speculation, we annealed CsPb₂Br₅ sample and measured the phase transition and meanwhile monitored the associated absorption and PL properties. Fig. 4a is the PXRD data of CsPb₂Br₅ annealed at different temperatures. As shown in this figure, the diffraction peaks corresponding to CsPbBr₃ and PbBr₂ were detected when the sample was treated with thermal annealing, and the intensity of the corresponding diffraction peaks increased with annealing temperature. It indicates that part of the CsPb₂Br₅ is decomposed into CsPbBr₃ as shown by eqn (4),

CsPb₂Br₅ → CsPbBr₃ + PbBr₂ (4)

Fig. 4 (a) PXRD pattern of tetragonal CsPb₂Br₅ particles after annealing at different temperatures (red squares represent CsPbBr₃ (PDF#18-0364), black dots represent PbBr₂), sample 1 is of high purity CsPb₂Br₅ (black) without annealing, sample 2 is CsPb₂Br₅ after annealing at 220 °C (red), sample 3 is CsPb₂Br₅ after annealing at 400 °C (blue). (b) XPS spectra of the pure as-grown CsPb₂Br₅ and annealed composite CsPb₂Br₅–CsPbBr₃ with element Pb 4f. (c) Low-resolution, and (d) high-resolution TEM image of a representative CsPb₂Br₅ cluster after annealing at 400 °C.

XPS characterization results in Fig. 4b reveals that there is no obvious peak shift of binding energy after the thermal annealing due to the same chemical valence of Pb in CsPb₂Br₅ and CsPbBr₃. The microstructure and composition of the annealed sample was further revealed by TEM and shown in Fig. 4c and d. The CsPb₂Br₅ and CsPbBr₃ are clearly observed in the high resolution TEM, which consistent with XRD results.

Fig. 5a shows the notable increase in absorption between 380 and 530 nm. This absorption band grows with increasing annealing temperature. When the sample was annealed at 400 °C, the absorption edge at 380 nm was not distinguishable any more due to being overwhelmed by the absorption of CsPbBr₃, which might explains the absence of the absorption edge at 380 nm as observed in previous reports. The consistency between the PXRD and absorption results demonstrates that the absorption at 515 nm is indeed induced by CsPbBr₃.

Fig. 5 (a), (b) Absorption and PL spectra of CsPb₂Br₅ particles. The black, red and blue represents the sample of as-grown, annealed at 220 °C, and 400 °C, respectively. The PL excitation is 400 nm. (c) DFT calculated band structure of CsPb₂Br₅ with a bandgap of 3.079 eV.

The PL spectra in Fig. 5b also shows the luminescence peaks corresponding to CsPbBr₃.^11,18 In addition, the position of PL peak shifts from 515 to 530 nm with thermal annealing. This is ascribed to the size effect. In the as-grown sample, the contents of CsPbBr₃ is too low to be detected by XRD and the grain size must be very small. Whereas, it could grown into larger size with the decomposition of CsPb₂Br₅ after thermal annealing. The band structure of CsPb₂Br₅ was investigated theoretically by first principles calculations based on DFT (Fig. 5c). It could be seen that CsPb₂Br₅ exhibits an indirect band gap of 3.079 eV with the conduction band minimum and valence band maximum at the Γ and K points respectively. The obtained bandgap value agrees reasonably well with the value as obtained from the absorption spectrum.

Moreover, we investigated the partial cation exchange of Cs by Rb to from Cs_1−xRb_xPb₂Br₅. When RbAc was introduced together with CsAc, Cs_1−xRb_xPb₂Br₅ compound was successfully synthesized. As shown in Fig. 6a, the ion exchange by Rb did not change the crystalline structure but the lattice constant was reduced due to the smaller atomic radius of Rb. Fig. 6b plots the absorption and PL spectra of RbPb₂Br₅. Compared to CsPb₂Br₅ in Fig. 3a, the absorption and PL are red-shifted, indicating that ion exchange of A-site could be utilized to tune the bandgap and extend the optical response of Cs_1−xRb_xPb₂Br₅ to the visible range.

Fig. 6 (a) PXRD patterns of the parent CsPb₂Br₅ particles and anion exchanged samples with increasing quantities of anion-exchange of Rb⁻. (b) Absorption (black) and PL spectra (red) of RbPb₂Br₅, the PL spectrum under 250 nm excitation.

4. Conclusions

In conclusion, we developed a novel environmentally green method to synthesize CsPb₂Br₅, where only water and ethanol were used as solvents. The optical property and electronic structure of CsPb₂Br₅ have been explored experimentally as well as theoretically. For the first time, we observed the ultraviolet emission from CsPb₂Br₅. Employing the technique of UV-vis absorption, PL, and first principle calculation, we find that the ultraviolet emission is ascribed to band-edge emission of CsPb₂Br₅, which has a bandgap of 3.13 eV. The visible emission at 515 nm that was previously ascribed to band edge emission of CsPb₂Br₅ is now unambiguously assigned CsPbBr₃, which emerged as a by-product due to thermally induced phase conversion between CsPb₂Br₅ and CsPbBr₃. This study not only clarifies the luminescence mechanism of CsPb₂Br₅ being under debate, but also provides valuable insights for the development of CsPb₂Br₅ based optoelectronic devices.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work has been supported by National Natural Science Foundation of China (No. 11174071, 11304088 and 51372180) and Special Technical Innovation Project of Hubei Province (No. 2016AAA035).

Notes and references

1. W. S. Yang, B. W. Park, E. H. Jung, N. J. Jeon, Y. C. Kim, D. U. Lee, S. S. Shin, J. Seo, E. K. Kim, J. H. Noh and S. I. Seok, Science, 2017, 356, 1376.
2. A. Kojima, K. Teshima, Y. Shirai and T. Miyasaka, J. Am. Chem. Soc., 2009, 131, 6050–6051.
3. J. B. You, L. Meng, T. B. Song, T. F. Guo, Y. Yang, W. H. Chang, Z. R. Hong, H. J. Chen, H. P. Zhou, Q. Chen, Y. S. Liu, N. D. Marco and Y. Yang, Nat. Nanotechnol., 2016, 11, 75.
4. Q. D. Tai, P. You, H. Q. Sang, Z. K. Liu, C. L. Hu, H. L. W. Chan and F. Yan, Nat. Commun., 2016, 7, 11105.
5. W. Z. Li, W. Zhang, S. V. Reenen, R. J. Sutton, J. D. Fan, A. A. Haghighirad, M. B. Johnston, L. D. Wang and H. J. Snaith, Energy Environ. Sci., 2016, 9, 490.
6. G. Niu, X. Guo and L. Wang, J. Am. Chem. Soc., 2015, 3, 8970–8980.
7. T. A. Berhe, W. N. Su, C. H. Chen, C. J. Pan, J. H. Cheng, H. M. Chen, M. C. Tsai, L. Y. Chen, A. A. Dubale and B. J. Hwang, Energy Environ. Sci., 2016, 9, 323–356.
8. N. Yantara, S. Bhaumik, F. Yan, D. Sabba, H. A. Dewi, N. Mathews, P. P. Boix, H. V. Demir and S. Mhaisalkar, J. Phys. Chem. Lett., 2015, 6, 4360.
9. S. W. Eaton, M. Lai, N. A. Gibson, A. B. Wong, L. Dou, J. Ma, L. W. Wang, S. R. Leone and P. D. Yang, Proc. Natl. Acad. Sci. U. S. A., 2016, 113, 1993.
10. J. Xue, Y. Gu, Q. S. Shan, Y. S. Zou, J. Z. Song, L. M. Xu, Y. H. Dong, J. H. Li and H. B. Zeng, Angew. Chem., Int. Ed., 2017, 56, 5232.
11. L. Protesescu, S. Yakunin, M. I. Bodnarchuk, F. Krieg, R. Caputo, C. H. Hendon, R. X. Yang, A. Walsh and M. V. Kovalenko, Nano Lett., 2015, 15, 3692.
12. K. H. Wang, L. Wu, L. Li, H. B. Yao, H. S. Qian and S. H. Yu, Angew. Chem., Int. Ed., 2016, 55, 8328.
13. X. S. Tang, Z. P. Hu, W. Yuan, W. Hu, H. B. Shao, D. J. Han, J. F. Zheng, J. Y. Hao, Z. G. Zang, J. Du, Y. X. Leng, L. Fang and M. Zhou, Adv. Opt. Mater., 2016, 5, 1600788.
14. X. L. Zhang, B. Xu, J. B. Zhang, Y. Gao, Y. J. Zheng, K. Wang and X. W. Sun, Adv. Funct. Mater., 2016, 26, 4595.
15. G. P. Li, H. Wang, Z. F. Zhu, Y. J. Chang, T. Zhang, Z. H. Song and Y. Jiang, Chem. Commun., 2016, 52, 11296.
16. J. Liang, C. X. Wang, Y. R. Wang, Z. R. Xu, Z. P. Lu, Y. Ma, H. F. Zhu, Y. Hu, C. C. Xiao, X. Yi, G. Y. Zhu, H. L. Lv, L. B. Ma, T. Chen, Z. X. Tie, Z. Jin and J. Liu, J. Am. Chem. Soc., 2016, 138, 15829–15832.
17. K. Nitsch, V. Hamplova, M. Nikl, C. Canale, K. Polak and M. Rodova, Chem. Phys. Lett., 1996, 258, 518–522.
18. X. M. Li, Y. Wu, S. L. Zhang, B. Cai, Y. Gu, J. Z. Song and H. B. Zeng, Adv. Funct. Mater., 2016, 26, 2435.

---

Footnote

† Electronic supplementary information (ESI) available: Full video of the chemical reaction; energy dispersive X-spectroscopy of CsPb₂Br₅; total energy calculation of unit cell for each reactant and the resultant chemical eqn (2). See DOI: 10.1039/c7ra10693a

---