One-dimensional lead-free perovskite single crystals with high X-ray response grown by liquid phase diffusion

Abstract

Organic–inorganic hybrid bismuth halides, as essential members of the lead-free family, have attracted much attention due to their unique properties. However, the growth of bismuth-based single crystals with potentially low defect density has encountered many challenges, such as intractable rapid crystallization, the use of toxic organic solvents, and halide oxidation caused by prolonged heating. Herein, we developed a liquid phase diffusion method to prepare large-sized one-dimensional hybrid bismuth single crystals. Adopting this method, centimeter-sized (1,5-pentanediamine)BiI₅ single crystals were harvested, which possess a hole trap density as low as 1.31 × 10⁹ cm⁻³ and a carrier mobility reaching 28.2 cm² V⁻¹ s⁻¹. The X-ray detectors based on (1,5-pentanediamine)BiI₅ single crystals exhibit a sensitivity value of 8120 μC Gy_air⁻¹ cm⁻² at −80 V. This work provided a new strategy for the growth of bismuth-based perovskite single crystals and inspired the growth of other hybrid perovskite materials.

---

Introduction

X-ray detectors are widely used in computed tomography, nondestructive product inspection, scientific research, etc.^1–3 At present, two detector technologies prevail: the indirect scintillator and the direct photoconductor.⁴ Between them, the semiconductor-based direct conversion form has attracted more attention as it outperformed the scintillators in the aspect of device configuration and detection sensitivity. Currently, the most advanced and popular direct X-ray imaging system is based on amorphous selenium (α-Se). However, its charge transport properties and X-ray stopping capacity are still far from satisfactory. Accordingly, candidates with high X-ray absorption capacity (e.g. PbI₂, CdZnTe, and HgI₂) have been extensively developed.⁵ Although the emergence of these candidates partially fulfilled the expectations, it also caused other challenges inherently. For instance, the growth of high-quality CdZnTe needs a processing temperature of 600 °C, far beyond the heat resistance limit of thin-film transistors.⁶ Recently, organic–inorganic hybrid lead halide perovskites have been recognized as a dark horse for the development of direct X-ray imaging due to their solution-processability, structural diversity, and excellent photophysical properties.^7–12 Detectors based on the perovskites demonstrated astounding performance, e.g. ultra-high sensitivity,¹³ an extremely low detection limit,¹⁴ and rapid light response.^15,16 Despite the abovementioned substantial advances, the use of lead has also raised environmental concerns.^17–20

Theoretically, the X-ray absorption coefficient of an element is proportional to its atomic number.²¹ Bismuth not only possesses a higher atomic number but also its trivalent ions share the same electronic configuration with Pb²⁺.^20,22–25 Therefore, bismuth-based hybrid halide perovskites, as a class of lead-free members, provides potential new players for high-performance detection materials as exemplified by (3-(aminomethyl)pyridine)BiI₅, (benzamidinium)BiI₄, MA₃Bi₂I₉, (CH₃NH₃CH₂CH₂NH₃CH₃)BiI₅ and (N-methyl-1,3-diaminopropanium)BiI₅.^17,26–30 In addition to enriching the bismuth-based perovskite family, researchers have also made a lot of attempts in crystal growth methods. For example, using the antisolvent vapor-assisted method, centimeter-sized 1D (pentamethylenediamine)BiBr₅ single crystals (SCs) were obtained.³¹ With the help of the top-seeded solution growth method, Luo's group prepared large-sized 1D (N-methyl-1,3-diaminopropanium)BiI₅.²⁹ Employing a temperature cooling method, (3-(aminomethyl)pyridine)BiI₅ as well as (CH₃NH₃CH₂CH₂NH₃CH₃)BiI₅ SCs were also harvested.^17,28 Additionally, the inverse temperature crystallization method was also adopted to prepare 0D MA₃Bi₂I₉ SCs from γ-butyrolactone solution.²⁷ Despite recent blooming advances, the aforementioned methods still suffer from some drawbacks. The top-seeded solution growth method, inverse temperature crystallization method, and temperature cooling method, usually require long periods of heating, which inevitably triggers I⁻ oxidation, greatly impairing crystal quality^32–35 while the antisolvent vapor method often needs toxic organic solvents such as toluene and dichloromethane. Moreover, the crystallization properties and kinetics of Bi-based perovskites are also significantly different from those of the lead ones, e.g. Bi-based perovskite precursors convert into target phase directly without mediating through crystalline solvated phases.^36–38 In these aspects, the crystallization of Bi-based perovskites should be comprehensively considered in the context of growth methods and their own features.

In this work, we first studied the solubility of Bi-based perovskite and its corresponding precursors and found that there is a huge discrepancy between them. Specifically, the solubility of the BiI₃ is more than 100 times that of the target product. Such a huge gap leads to the precipitation of a large number of microcrystals. Based on this understanding, we propose a liquid phase diffusion (LPD) growth strategy to control the precipitation amount of crystal nucleus, which allows harvesting of centimeter-sized single crystals (SCs). X-ray detectors based on the prepared (1,5-PDA)BiI₅ (1,5-PDA is 1,5-pentanediamine) SCs deliver a sensitivity value of 8120 μC Gy_air⁻¹ cm⁻² under −80 V. Our finding provides new insight into SC growth, which expands the crystal growth spectrum of mixed perovskite materials. This strategy can also be easily extended to other perovskite-based materials to improve their optoelectronic performance.

Results and discussion

Crystal structure of (1,5-PDA)BiI₅

We first use the popular temperature cooling method to grow (1,5-PDA)BiI₅ SCs (the detailed method can be seen in the ESI†). At room temperature, (1,5-PDA)BiI₅ crystallizes in the orthorhombic space group of P2₁2₁2₁ (point group of 222) with cell parameters: a = 8.5824(4) Å, b = 14.7005(7) Å, c = 14.1760(7) Å, and V = 1788.52(15) ų, of which asymmetric unit consists of one BiI₅⁻ anion and one protonated 1,5-PDA cation (Table S1, ESI†). Each Bi³⁺ metal cation is coordinated with six I⁻ anions to form a BiI₆ octahedron, in which four I⁻ anions are terminal and two act as a bridge to connect two neighboring BiI₆ octahedrons, showing that 1D corner-shared (BiI₅)_n²ⁿ⁻ zigzag chains extend infinitely along a-axis direction, as shown in Fig. 1. The Bi–I bond distances range from 2.8928(11) to 3.3144(10) Å; I–Bi–I bond angles range from 83.990(13) to 97.90(3)°, Table S2 and S3 (ESI†). Δd (distortion index) and σ² (bond angle variance) were employed to evaluate the distortion level of the BiI₆ octahedrons, respectively (the detailed calculation process can be seen in the ESI†). Δd and σ² were separately calculated to be 0.03909 and 18.55°,² indicating significant distortion. The 1,5-PDA organic cations occupied the cavities that can be seen to be composed of the four (BiI₅)_n²ⁿ⁻ zigzag chains (Fig. S1, ESI†). As shown in Fig. S2a (ESI†), the Hirshfeld surface mapped with d_norm (normalized contact distance) disclose that the main interaction between organic cations and inorganic formwork is N–H⋯I Hydrogen bonds range from 2.766 to 2.875 Å. Fingerprint plots suggest N–H⋯I Hydrogen is 75.3% of the total Hirshfeld surface as shown in Fig. S2b (ESI†).

Fig. 1 The packing structure of (1,5-PDA)BiI₅.

Growth and optimization of (1,5-PDA)BiI₅ SCs

During the aforementioned temperature cooling process, we found that the colour of the solution turned into black from pale yellow, suggesting severe oxidation of iodide ions after a long period of continuous heating.³⁹ Moreover, we found it difficult to control the number of precipitated nuclei as evidenced by a large amount of microcrystals were obtained at the bottom of the flask. Further experiments show that there is a huge difference in solubility between the precursors and target products. By contrast to the good solubility of 1,5-PDAI₂ (2.14 mmol ml⁻¹) and BiI₅ (3.88 mmol ml⁻¹) in hydroiodic acid, the solubility of the (1,5-PDA)BiI₅ is as low as 0.0348 mmol ml⁻¹, even if the hydroiodic acid is heated to near boiling point (127 °C), the solubility of the (1,5-PDA)BiI₅ can only reach 0.0875 mmol ml⁻¹ (Fig. 2a). Due to such a low solubility, a large number of (1,5-PDA)BiI₅ crystallites formed rapidly once the precursors mixed together, hindering further growth of single-crystals. The process of crystallization can be divided into two steps: nucleation and growth.³ Clearly, in our case, the key factor for growing large-sized and high-quality SCs is how to effectively control the number of nuclei. Based on the above analysis, we developed the LPD method to grow (1,5-PDA)BiI₅ SCs. Briefly, it is carried out in an H-shaped test tube composed of three parts (A, B, and C), as shown in Fig. 2b. First, hydroiodic acid containing BiI₅ and 1,5-PDAI₂ were added into test tubes A and C, respectively. Then hydroiodic acid acting as a buffer to prevent the formation of a large number of (1,5-PDA)BiI₅ crystallites was carefully injected into test tube B. As time goes on, precursor 1,5-PDAI₂ in test tube A and BiI₅ in test tube C would continually diffuse into the middle of test tube B. When the concentration of (1,5-PDA)BiI₅ in the middle of test tube B reaches supersaturation, crystal nuclei would be formed. The formation of crystal nuclei would decrease the concentration of BiI₅ and 1,5-PDAI₂ around crystal nuclei, which brings about the concentration gradient. Therefore, 1,5-PDAI₂ and BiI₅ would resume diffusion toward crystal nuclei and crystal nuclei would continually grow. Notably, the diffusion of precursors and crystal growth are synchronous. After a month, black crystals were obtained in the middle of the test tube B.

Fig. 2 (a) Solubility of precursors and target product. (b) The schematic diagram for (1,5-PDA)BiI₅ large-sized single-crystal growth using the LPD method. (c) Comparison between experimental PXRD pattern and calculated counterpart. Inset: Optical image of the SC of (1,5-PDA)BiI₅. (d) TG measurement for (1,5-PDA)BiI₅.

The digital photograph of the as-prepared (1,5-PDA)BiI₅ SC is presented in the inset of Fig. 2c. The powder X-ray diffraction measurements verified the phase purity of (1,5-PDA)BiI₅ (Fig. 2c). Thermogravimetric (TG) measurement was adopted to evaluate the thermal stability. As shown in Fig. 2d, there is no obvious mass loss until 573 K, suggesting the decomposition of (1,5-PDA)BiI₅ greater than this temperature.

The UV-vis absorption spectrum of (1,5-PDA)BiI₅ was recorded for evaluation of the potential of (1,5-PDA)BiI₅ in photoelectronic applications. As shown in Fig. 3a, there is an absorption edge at about 650 nm. By fitting variant Tauc equation (the detailed process can be seen in ESI†), the optical band gap of (1,5-PDA)BiI₅ was determined to be 1.95 eV, which is close to MA₃Bi₂I₉ of 1.99 eV,²⁷ and greater than (3-(aminomethyl)pyridine)BiI₅ of 1.74 eV,¹⁷ (benzamidinium)BiI₄ of 1.77 eV,²⁶ (N-methyl-1,3-diaminopropanium)BiI₅ of 1.83 eV,²⁹ and (CH₃NH₃CH₂CH₂NH₃CH₃)BiI₅ of 1.86 eV.²⁸ The theoretical calculations were further adopted to disclose the band gap characteristic of (1,5-PDA)BiI₅. As shown in Fig. 3b, the bottom of the conduction band (CBM) and the top of the valence band (VBM) are both located at the G point, indicating a direct band gap similar to (CH₃NH₃CH₂CH₂NH₃CH₃)BiI₅.²⁸ The difference value between the CBM and the VBM, the calculated band gap, is approximately 2 eV, on par with that of the experimental result. The partial densities of states (PDOS) calculations were carried out to investigate the origin of the band gap of (1,5-PDA)BiI₅. As shown in Fig. S3 (ESI†), the VBM mainly consists of I 5p orbitals with a small number of Bi 6s and 6p orbitals, while CBM originates from Bi 6p and I 5p orbitals.

Fig. 3 (a) UV-vis absorption spectra of (1,5-PDA)BiI₅. Inset shows the Tauc plot of (1,5-PDA)BiI₅. (b) Band structure of (1,5-PDA)BiI₅. (c) Current–voltage curve of the (1,5-PDA)BiI₅ SC device.

We further studied the trap density (n_trap) and carrier mobility (μ) of the prepared crystal based on the space charge-limited current (SCLC) method. As shown in Fig. 3c, the ohmic region, trap-filling region and the trap-free SCLC region can be identified clearly from the current–voltage curve. The hole trap density can be obtained from the following equation:

where ε₀ is the vacuum permittivity, ε is the relative dielectric constant, n_trap is the trap density, and L is the thickness of SCs. The as-prepared crystal shows a trap density of 1.31 × 10⁹ cm⁻³, which is lower than that of the previously reported Bi-based perovskite SCs, illustrating the LPD method is an effective way to grow high-quality SCs (Table S4, ESI†). We also calculated the carrier mobility based on the Child's region by the Mott–Gurney law.⁶ The carrier mobility is 28.207 cm² V⁻¹ s⁻¹, which is higher than those of most Bi-based perovskite materials.

Device performance of (1,5-PDA)BiI₅ SC X-ray detectors

Inspired by the prepared high-quality crystals, we then try to evaluate their performance as X-ray detectors. We first calculated the X-ray absorption coefficients of (1,5-PDA)BiI₅ and several common traditional semiconductors by the photon cross-section database.¹ As shown in Fig. 4a, the absorption coefficient of (1,5-PDA)BiI₅ is much higher than that of the semiconductor Si and is close to that of the α-Se and CdTe within the photon energy range of 5 keV to 100 MeV. This is consistent with a fact that Bi possesses a higher atomic number and indicates its X-ray absorption capacity is similar to or even better than those of traditional semiconductors. Fig. 4b shows the relationship between the thickness and the attenuation efficiency of (1,5-PDA)BiI₅ and several traditional detector materials at 40 keV X-ray photons. The attenuation efficiency of (1,5-PDA)BiI₅ and several traditional semiconductor materials increases with the increase of thickness under irradiation. At the same thickness, the attenuation efficiency of (1,5-PDA)BiI₅ is obviously higher than that of the Si and α-Se. Taking 0.5 mm as an example, (1,5-PDA)BiI₅ attenuates 95% of the incident photons, while that of the Si and α-Se is only about 8% and 82%, respectively.

Fig. 4 Performance of the (1,5-PDA)BiI₅ SC X-ray detector. (a) Absorption coefficients of (1,5-PDA)BiI₅, Si, α-Se, and CdTe as a function of photon energy. (b) Attenuation efficiency of (1,5-PDA)BiI₅, Si, α-Se, and CdTe for 40 keV X-ray photons versus thickness. (c) Bias-dependent photoconductivity of the (1,5-PDA)BiI₅ device. Inset: Schematic illustration of the device used for photoconductivity measurement. (d) Relationship between time and current density of (1,5-PDA)BiI₅, (1,5-PDA)BiI₅ device under −5 V bias voltage. (e) X-Ray response current at various dose rates (the sensitivity can be derived from the slope of the fitting line). (f) Representative photo-response curve under low dose rate radiation with a (1,5-PDA)BiI₅ bias of −4 V. (g) Signal-to-noise ratio of the detector under low dose rate radiation with different bias voltage. (h) Device stability upon continuous radiation with a dose rate of 113.1 μGy_air s⁻¹ under −10 V.

After undergoing a range of interactions between X-ray and detection materials, a massive amount of carriers were generated, whose lifetime and mobility product (μτ) determined the mean distance drifted under a unit electric field before recombination. This value was obtained by fitting the photoconductivity curve (Fig. 4c) based on the modified Hecht equation.² The μτ product of the (1,5-PDA)BiI₅ SC is 7.49 × 10⁻³ cm² V⁻¹, which is comparable to that of the CdZnTe (9.13 × 10⁻³ cm² V⁻¹)³ and is just slightly lower than that of the MAPbBr₃ (1.43 × 10⁻² cm² V⁻¹).⁴⁰ In order to study the detection performance of the device, we designed a parallel planar electrode, which consists of a 120 nm-thick Au. The detector sensitivity is an important parameter of advanced imaging, which is defined as the charge collected per unit area of radiation exposure (μC Gy_air⁻¹ cm⁻²). Fig. 4d shows the representative light response feature of the as-prepared detector under different dose rates (for other photoelectric response curves please see Fig. S4, ESI†). It displays a –42 pA cm⁻² dark current without obvious drift during test that lasted more than ten minutes. Moreover, the on–off switching feature can also be clearly identified under low radiation dose rates, indicating excellent light response properties. The corresponding sensitivity (S) value can be obtained according to the following formula:

where I_p and I_d are the photocurrent and dark current, respectively, D is the irradiation dose rate, and A is the effective area of the device.⁴ The photocurrent density at different dose rates shows good linearity. The corresponding S was listed in Fig. 4e. Under high bias voltage of −80 V, the detector shows an impressive value of 8120 μC Gy_air⁻¹ cm⁻² and even for the low bias voltage, it also delivers a 3859 μC Gy_air⁻¹ cm⁻² sensitivity value, much higher than those of the lead-free perovskites, such as MA₃Bi₂I₉ (1947 μC Gy_air⁻¹ cm⁻²), (BA)₂CsAgBiBr₇ (4.2 μC Gy_air⁻¹ cm⁻²), (chloropropylammonium)₄AgBiBr₈ (0.8 μC Gy_air⁻¹ cm⁻²), and (4,4-difluoropiperidinium)₄AgSbI₈ (704.8 μC Gy_air⁻¹ cm⁻²).^23–25

Encouraged by the aforementioned outstanding weak light response feature, we further studied the detection limit of the detector. IUPAC defines the detection limit as the equivalent dose rate that generates a signal greater than three times the noise level under a given electric field. The calculation formula for the signal-to-noise ratio is

where the signal current (I_signal) was derived by subtracting the average photocurrent (Ī_photo) by the average dark current (Ī_dark) while the noise current (I_noise) was obtained by calculating the standard deviation of the photocurrent. As shown in Fig. 4f, the detector shows a very clear switching characteristic curve under a dose rate of 4.5 nGy_air s⁻¹ and 9.5 nGy_air s⁻¹. Fig. 4g shows the signal-to-noise ratio (SNR) of the detector under different bias voltages (Fig. S5, ESI†). Obviously, the signal-to-noise ratio of the detector is greater than 3, indicating that the detector has a very low detection limit. We also tested the long-term stability of the device. After 45 minutes of irradiation (dose rate of 113.1 μGy_air s⁻¹), the detector delivered a stable photocurrent density of −129 nA cm⁻² without obvious drift (Fig. 4h), illustrating excellent stability.

Conclusions

In summary, the LPD method is demonstrated to control the precipitation amount of crystal nucleus which enables us to harvest centimeter-sized SCs in perovskite systems with huge solubility discrepancy. As a result, the harvested (1,5-PDA)BiI₅ SCs show a low trap-state density of 1.31 × 10⁹ cm⁻³ and a high carrier mobility of 28.207 cm² V⁻¹ s^-1. X-ray detectors based on the prepared SCs not only delivered an impressive sensitivity value (8120 μC Gy_air⁻¹ cm⁻²), but also exhibited an ultra-low detection limit (4.5 nGy_air s⁻¹). We believe that this simple and feasible crystal growth method could be applied to the preparation of other materials.

Author contributions

C.-F. Wang designed the single crystal growth method, prepared materials, determined the single crystal structure, carried out DFT calculations and wrote the original draft; C.-L. Fang prepared the devices, performed device performance measurements and wrote the original draft; Y. Feng prepared synthetic materials; S.-Y. Liu performed device performance measurements; Y. Zhang, H.-Y Ye and L.-P. Miao provided suggestions for the research. L. Liu wrote and revised draft. All authors have given approval to the final version of the manuscript.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The authors acknowledge the support from the National Natural Science Foundation of China (22109057 and 21875093), the Natural Science Foundation of Jiangxi Province (20212BAB214021, 20204BCJ22015, and 20202ACBL203001), and the Science and Technology Project of Jiangxi Provincial Department of Education (GJJ200836 and GJJ210847).

Notes and references

1. M. Ahmadi, T. Wu and B. Hu, Adv. Mater., 2017, 29, 1605242.
2. Y. Zhang, Y. Liu, Z. Xu, H. Ye, Z. Yang, J. You, M. Liu, Y. He, M. G. Kanatzidis and S. F. Liu, Nat. Commun., 2020, 11, 2304.
3. Y. Liu, H. Ye, Y. Zhang, K. Zhao, Z. Yang, Y. Yuan, H. Wu, G. Zhao, Z. Yang, J. Tang, Z. Xu and S. Liu, Matter, 2019, 1, 465–480.
4. Y. C. Kim, K. H. Kim, D.-Y. Son, D.-N. Jeong, J.-Y. Seo, Y. S. Choi, I. T. Han, S. Y. Lee and N.-G. Park, Nature, 2017, 550, 87–91.
5. H. Wei and J. Huang, Nat. Commun., 2019, 10, 1066.
6. S. Kasap, J. B. Frey, G. Belev, O. Tousignant, H. Mani, J. Greenspan, L. Laperriere, O. Bubon, A. Reznik, G. DeCrescenzo, K. S. Karim and J. A. Rowlands, Sensors, 2011, 11, 5112–5157.
7. X. Xu, W. Qian, S. Xiao, J. Wang, S. Zheng and S. Yang, EcoMat, 2020, 2, e12064.
8. L. Mao, C. C. Stoumpos and M. G. Kanatzidis, J. Am. Chem. Soc., 2018, 141, 1171–1190.
9. H. Hu, G. Niu, Z. Zheng, L. Xu, L. Liu and J. Tang, EcoMat, 2022, 4, e12258.
10. Y. Zhou, J. Chen, O. M. Bakr and O. F. Mohammed, ACS Energy Lett., 2021, 6, 739–768.
11. Z. Li, F. Zhou, H. Yao, Z. Ci, Z. Yang and Z. Jin, Mater. Today, 2021, 48, 155–175.
12. Y. Y. Chen, C. H. Gao, T. Yang, W. J. Li, H. J. Xu and Z. H. Sun, Chin. J. Struct. Chem., 2022, 41, 2204001–22040011.
13. Y. Song, L. Li, M. Hao, W. Bi, A. Wang, Y. Kang, H. Li, X. Li, Y. Fang, D. Yang and Q. Dong, Adv. Mater., 2021, 33, 2103078.
14. J. Peng, Y. Xu, F. Yao, H. Huang, R. Li and Q. Lin, Matter, 2022, 5, 2251–2264.
15. H. Li, J. Song, W. Pan, D. Xu, W. A. Zhu, H. Wei and B. Yang, Adv. Mater., 2020, 32, 2003790.
16. R. Zhuang, X. Wang, W. Ma, Y. Wu, X. Chen, L. Tang, H. Zhu, J. Liu, L. Wu, W. Zhou, X. Liu and Y. Yang, Nat. Photonics, 2019, 13, 602–608.
17. X. Yang, Y. H. Huang, X. D. Wang, W. G. Li and D. B. Kuang, Angew. Chem., Int. Ed., 2022, e202204663,  DOI:10.1002/anie.202204663.
18. X. Liu, H. Li, Q. Cui, S. Wang, C. Ma, N. Li, N. Bu, T. Yang, X. Song, Y. Liu, Z. Yang, K. Zhao and S. F. Liu, Angew. Chem., Int. Ed., 2022, e202209320,  DOI:10.1002/anie.202209320.
19. Z. Xu, H. Wu, D. Li, W. Wu, L. Li and J. Luo, J. Mater. Chem. C, 2021, 9, 13157–13161.
20. C. F. Wang, H. Li, M. G. Li, Y. Cui, X. Song, Q. W. Wang, J. Y. Jiang, M. M. Hua, Q. Xu, K. Zhao, H. Y. Ye and Y. Zhang, Adv. Funct. Mater., 2021, 31, 2009457.
21. X. Xu, W. Qian, S. Xiao, J. Wang, S. Zheng and S. Yang, EcoMat, 2020, 2, e12064.
22. X. G. Zhao, D. W. Yang, J. C. Ren, Y. H. Sun, Z. W. Xiao and L. J. Zhang, Joule, 2018, 2, 1662–1673.
23. C. F. Wang, H. Li, Q. Ji, C. Ma, L. Liu, H. Y. Ye, B. Cao, G. Yuan, H. F. Lu, D. W. Fu, M. G. Ju, J. Wang, K. Zhao and Y. Zhang, Adv. Funct. Mater., 2022, 32, 2205918.
24. W. Guo, X. Liu, S. Han, Y. Liu, Z. Xu, M. Hong, J. Luo and Z. Sun, Angew. Chem., Int. Ed., 2020, 59, 13879–13884.
25. Z. Xu, X. Liu, Y. Li, X. Liu, T. Yang, C. Ji, S. Han, Y. Xu, J. Luo and Z. Sun, Angew. Chem., Int. Ed., 2019, 58, 15757–15761.
26. C. Ma, H. Li, M. Chen, Y. Liu, K. Zhao and S. Liu, Adv. Funct. Mater., 2022, 32, 2202160.
27. Y. Liu, Z. Xu, Z. Yang, Y. Zhang, J. Cui, Y. He, H. Ye, K. Zhao, H. Sun, R. Lu, M. Liu, M. G. Kanatzidis and S. Liu, Matter, 2020, 3, 180–196.
28. L. Yao, G. Niu, L. Yin, X. Du, Y. Lin, X. Den, J. Zhang and J. Tang, J. Mater. Chem. C, 2020, 8, 1239–1243.
29. K. Tao, Y. Li, C. Ji, X. Liu, Z. Wu, S. Han, Z. Sun and J. Luo, Chem. Mater., 2019, 31, 5927–5932.
30. X. Zheng, W. Zhao, P. Wang, H. Tan, M. I. Saidaminov, S. Tie, L. Chen, Y. Peng, J. Long and W.-H. Zhang, J. Energy Chem., 2020, 49, 299–306.
31. Y. Xu, J. Hu, X. Du, X. Xiao, M. Li, J. Tang, J. Chen and Y. He, Chem. – Asian J., 2021, 16, 4137–4144.
32. B. Kou, W. Zhang, C. Ji, Z. Wu, S. Zhang, X. Liu and J. Luo, Chem. Commun., 2019, 55, 14174–14177.
33. 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, 1376–1379.
34. W. Zhang, B. Kou, Y. Peng, Z. Wu, Y. Yao, D. Dey, L. Li and J. Luo, J. Mater. Chem. C, 2018, 6, 12170–12174.
35. W. Zhang, X. Liu, L. Li, Z. Sun, S. Han, Z. Wu and J. Luo, Chem. Mater., 2018, 30, 4081–4088.
36. M.-C. Tang, D. Barrit, R. Munir, R. Li, J. M. Barbé, D.-M. Smilgies, S. Del Gobbo, T. D. Anthopoulos and A. Amassian, Solar RRL, 2019, 3, 1800305.
37. Q. Jia, C. Li, W. Tian, M. B. Johansson, E. M. J. Johansson and R. Yang, ACS Appl. Mater. Interfaces, 2020, 12, 43876–43884.
38. T. Singh, A. Kulkarni, M. Ikegami and T. Miyasaka, ACS Appl. Mater. Interfaces, 2016, 8, 14542–14547.
39. Q. Lv, Z. Lian, Q. Li, J. L. Sun and Q. Yan, Chem. Commun., 2018, 54, 1049–1052.
40. W. Wei, Y. Zhang, Q. Xu, H. Wei, Y. Fang, Q. Wang, Y. Deng, T. Li, A. Gruverman, L. Cao and J. Huang, Nat. Photonics, 2017, 11, 315–321.

---

Footnotes

† Electronic supplementary information (ESI) available. CCDC 2207566. For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d2tc04481a

‡ These authors contributed equally to this work.

---