High-performance perovskite solar cells fabricated by vapor deposition with optimized PbI₂ precursor films

Abstract

The quality of a perovskite film determines the performance of a perovskite solar cell. A novel hybrid physical-chemical vapor deposition (HPCVD) method is presented to grow high-quality CH₃NH₃PbI₃ films. These films were synthesized in a vacuum quartz tube with a constant growth temperature of 100 °C, resulting in the uniform film with grain sizes up to 800 nm and surface roughness of about RMS 17.4 nm. Combined with an optimized spin-coating process for PbI₂ precursor films, a high-performance CH₃NH₃PbI₃ solar cell's power conversion efficiency (PCE) can reach up to 14.2%. When treated in a controlled harsh environment at 80 °C for 96 hours, ten solar cells maintained 78% of their initial efficiency on average, which demonstrates the effectiveness of this HPCVD method.

---

Introduction

Hybrid inorganic and organic perovskite materials have shown tremendous potential in photovoltaic research and development in recent years.^1,2 Since the first perovskite solar cell was introduced in 2009, its power conversion efficiency (PCE) has jumped from 3.8% (ref. 3) to 20.1% (ref. 4) in five years, unprecedented in the history of photovoltaics (PV). In the meantime, more mature silicon-based PV research has shifted to the design and construction of nanostructures on solar cell surfaces to collect light more efficiently through concentration and trapping.^5–7 A typical perovskite solar cell (e.g., CH₃NH₃PbI₃) can be prepared by simple solution-based methods with such advantages as high power conversion efficiency and low material and fabrication cost.^8–11 Due to the superior properties of perovskite materials as optical absorption layers for a solar cell, such as an appropriate direct band gap,¹² high absorption length in the visible light spectrum,² long carrier diffusion length and intrinsically low defect levels,^13–15 perovskite solar cells have become the most competitive research area in photovoltaics in recent years. The crystalline quality of a perovskite film, including perovskite crystalline grains and the uniformity of morphology, greatly determines the performance of perovskite solar cells.^11,16–18 In addition, optical,^19,20 electrical^21–24 and semiconductor properties^13,14,25,26 of perovskite materials are greatly determined by synthesis methods. For broad application of perovskite solar cells in the future, it is necessary to develop methods to synthesize high-quality, uniform, and stable perovskite film at a large scale.

Hybrid perovskite was synthesized by the chemical reaction of two precursors, lead halide (e.g., PbI₂) and methylammonium halide (e.g., CH₃NH₃I). Solution-based and vapor-based methods are the two most popular methods utilized to synthesize perovskite films. The solution-based method for perovskite solar cells includes one-step and two-step deposition methods. In the one-step deposition method,⁸ lead halide (e.g., PbI₂) and methylammonium halide (e.g., CH₃NH₃I) are mixed in an organic solvent, such as N,N-dimethylformamide (DMF), N-methyl-2-pyrrolidone γ-butyrolactone (GBL), or dimethylsulphoxide (DMSO), followed by spin coating of the mixture solution and baking on a substrate to form a perovskite thin film. Meanwhile, perovskite thin films can also be prepared by two-step deposition.⁹ One major challenge of solution-based methods is that the materials used to make perovskite in solution crystallize too fast to form a uniform perovskite film due to the uncontrollable and slow evaporation speed of the perovskite solvent.¹¹ Recently, various procedures of perovskite synthesis have been demonstrated to improve the crystallization process for solution-based methods: (1) dripping toluene when the perovskite solution is spin coated,¹¹ and (2) adding argon flow when the solution is spin coated onto the substrate to enhance the solvent evaporation process to acquire perovskite films with smooth surface, uniform thickness, and large grain size.^27,28 Additionally, post-thermal annealing the perovskite films after spin coating in ambient air with 30% ± 5% relative humidity was introduced to enhance the re-crystallization effect for forming uniform films and reducing the boundary defects between perovskite grains.^17,29 Nonetheless, it is still technically demanding to synthesize uniform perovskite film of high quality with small surface roughness using solution-based methods.

Several vapor-based methods have also been developed to grow high-quality perovskite film, including dual-source co-evaporation,^16,30 vapor assistant solution deposition,¹⁸ and hybrid chemical vapor deposition (CVD).^31–34 In dual-source co-evaporation, methylammonium halide and lead halide precursors were evaporated at 120 °C and 325 °C simultaneously in an ultra-high vacuum chamber and reacted on a fluorine-doped tin oxide (FTO) glass substrate to fabricate a solar cell with 15.4% efficiency.¹⁶ In the vapor assistant solution deposition process (VASP), a layer of PbI₂ film was formed by spin coating firstly, and then reacting with methylammonium halide (e.g., CH₃NH₃I) vapor in a glove box to generate a uniform perovskite film. A solar cell of 12.1% efficiency was achieved with VASP perovskite films, with crystalline grain sizes and surface roughness of about 500 nm and RMS 23.2 nm, respectively.¹⁸ By a chemical vapor deposition method, single crystalline perovskite nanoplateles were grown on a muscovite mica surface with an electron diffusion length greater than 200 nm.³¹ Another vapor-based method using low-pressure chemical vapor deposition (LPCVD) has been applied to synthesize perovskite film, which was utilized to make a solar cell with an efficiency of 12.74%.³⁴ However, vapor-based methods are not widely used to synthesize high-quality perovskite thin films due to the complexity and challenge of controlling chemical and physical reactions in vapor phase. Perovskite films synthesized using dual-source co-evaporation needed to deal with two solid precursors at two different melting temperatures: one is lead halide, with a melting temperature at 325 °C, and the other is methylammonium halide, with a melting temperature at 125 °C. From previous research findings, perovskite thin film would degenerate over 145 °C.³⁵ Therefore, traditional evaporation and CVD process would not be appropriate to synthesize perovskite films without optimizing the conventional CVD system and deposition process. Meanwhile, critical parameters for the vapor reaction process, for instance reaction temperature and vapor pressure, could not be optimized for these vapor based processes to make high-quality perovskite materials.

In this paper, we introduce a novel HPCVD method to synthesize high-quality perovskite films of CH₃NH₃PbI₃ for fabricating high-performance solar cells. Compared to existing solution and vapor based methods, CH₃NH₃PbI₃ films are synthesized in a vacuum and isothermal environment through a uniform gas phase and solid phase reaction. In our previous work, several parameters for the HPCVD method were discussed to optimize the perovskite film synthesis, such as reaction temperature³⁶ and vapor pressure.³⁷ With a high-quality PbI₂ precursor film synthesized by an optimized spin-coating process, the quality of CH₃NH₃PbI₃ thin films could be further improved. The reaction temperature of 100 °C is one of the lowest synthesis temperatures for CH₃NH₃PbI₃ material, as far as we know. In ambient environment, PbI₂ solid thin films were fabricated by an optimized solution-based spin-coating process on a mesoporous TiO₂ (m-TiO₂)/compact TiO₂ (c-TiO₂)/FTO substrate. Next, purified CH₃NH₃I crystal solid precursor was laid into a quartz boat, which was loaded into an isothermal vacuum quartz tube, and the evaporated CH₃NH₃I vapor would react with the solid PbI₂ film to form a uniform CH₃NH₃PbI₃ thin film. However, a different CVD method has been reported³¹ to synthesize perovskite films by utilizing nitrogen flow to bring the CH₃NH₃I vapor, evaporating from a high-temperature zone (185 °C), to react with PbI₂ films at a relatively low temperature zone (about 130 °C). In our method, a vacuum of 2 mTorr and a low constant temperature of 100 °C were maintained to reduce the perovskite film growth rate and decrease perovskite film defects. A high PCE of up to 14.2% (reverse scanning) was achieved for these solar cells. After treatment in a controlled harsh environment (80 °C for 96 hours), ten solar cells produced by HPCVD maintained 78% of their initial efficiency (decreasing from 8.9% to 6.9%, on average), which demonstrates the effectiveness of this new HPCVD method. This HPCVD method is comparable with the traditional semiconductor material growth methods, and high-quality perovskite films could be synthesized under precise process controls. In future, high-performance perovskite solar cell could have great potential to be produced on a large scale at low cost by HPCVD for various application scenarios.

Experimental

Substrate preparation

FTO transparent conductive glass (Tec15, Pilkington) was patterned by laser etching. The substrates were cleaned in deionized (DI) water, ethanol, acetone, and isopropyl alchohol (IPA) sequentially for 15 min by ultrasonication. Then, fifteen minutes of oxygen plasma treatment was used to eliminate organic residues on the surfaces of the substrates.

Solar cell fabrication

A 20–40 nm-thick compact TiO₂ layer was spin coated on the FTO substrate according to the previous paper.¹⁶ Then, a 200 nm-thick mesoporous TiO₂ layer was spin coated onto the sintered compact TiO₂ surface at 5000 rpm for 15 s using a commercial TiO₂ paste (NJU-SR, Sunlaite, Suzhou, China) diluted in ethanol at 20% weight ratio. The mesoporous TiO₂ films were formed after sintering at 500 °C for 30 min to remove organics in the paste. Then, the TiO₂-coated substrates were treated in a 50 mM TiCl₄ aqueous solution at 70 °C for 30 min. After cleaning with DI water and ethanol, the substrates were sintered again at 500 °C for 30 min. PbI₂/N,N-dimethylformamide (DMF, Sigma) solution (461 mg mL⁻¹) was prepared on a hot plate at 70 °C with stirring. PbI₂/DMF solution was maintained at 70 °C before spin coating on substrates to keep its viscosity at an optimized range. After heated at 70 °C for 30 min to thoroughly remove the DMF residue, the PbI₂ substrates were loaded into a quartz tube evenly spread with CH₃NH₃I powder. After the tube was sealed with flanges and o-rings and vacuumized with a oil pump, 80 sccm nitrogen flow was purged into the tube for 20 min, and the tube was heated simultaneously to 50 °C to remove residual water vapor and oxygen gas inside. Then, the nitrogen flow was cut off, and the quartz tube was pumped to a vacuum level of 2 mTorr. The exact temperature in the tube was 100 °C, though the heating program of the quartz tube furnace was set at 110 °C, which was calibrated by a wireless PT1000 temperature recorder (TP-1000-W1, A-Volt Co., Ltd, Beijing, China) placed inside the vacuum tube during the whole heating procedure (calibrated temperature curve shown in Fig. S2†). After the solid PbI₂ films and CH₃NH₃I gas reacted in the tube for three hours, the substrates with perovskite films were removed from the tube and annealed at 100 °C for 10 min at a relative humidity of 30% ± 5% in ambient air. Then, hole transport layers were formed by spin coating 150 nm spiro-OMeTAD (Lumtech, Taiwan) on top of the perovskite films following previously reported methods.⁹ Finally, 60 nm gold films were deposited by E-beam evaporation as back electrodes at a spread of 0.3 angstrom per second. Dimension of the gold electrode pad was 4 × 4 mm.

Characterizations

Surface profiles of PbI₂ films, perovskite films, and the perovskite solar cell cross-section were measured by a field-emission scanning electron microscope (FE-SEM, Zeiss Merlin). Current density–voltage (J–V) curves of solar cells were characterized by a Semiconductor Device Parameter Analyzer (B1500A, Keysight, USA). The J–V curves were measured under AM 1.5 illumination (100 mW cm⁻²) generated by a Newport ABB (94021A) solar simulator in ambient air. Light intensity of the solar simulator was calibrated by a Newport mono-silicon reference cell (Newport calibration cert. #0702). A 3 × 3 mm steel mask covered the solar cells during the measurements. The voltage scan condition of J–V curves was 1.2 V to −0.1 V to 1.2 V with a 20 mV step and 50 ms delay. Also, incident photon-to-current conversion efficiencies (IPCE) of the perovskite solar cells produced by HPCVD were measured in the wavelength range of 300 to 900 nm (Oriel IQE 200, Newport, USA) at direct current (DC) mode.

Results and discussions

The high quality of the CH₃NH₃PbI₃ film could be attributed to the advantages of the HPCVD method, including the controllable vacuum environment, isothermal growing conditions, as well as the optimized PbI₂ solid precursor film synthesis and the isothermal process prior to and after the perovskite film synthesis.

Vacuum level and constant growing temperature are two critical parameters in achieving high-quality CH₃NH₃PbI₃ films. Typically, glove boxes filled with nitrogen were used to synthesize perovskite materials by solution methods to prevent solar cell degradation caused by high humidity and oxygen in ambient air.⁹ For our method, the quartz tube was inlet with 80 sccm nitrogen flow and pre-heated at 50 °C in order to remove probable water vapor and oxygen gas trapped inside these two precursor materials. After turning off the nitrogen flow, the quartz tube was pumped to 2 mTorr. Such vacuum level of the tube was maintained by running the pump consistently to avoid leakage contamination from the ambient environment. Two solid precursors (PbI₂ and CH₃NH₃I) were placed inside an isothermal region with a length of about 50 cm. Once the tube temperature was elevated to a constant temperature of 100 °C, CH₃NH₃I vapor was generated and reacted with the solid PbI₂ precursor thin films. High-quality CH₃NH₃PbI₃ films were achieved by this vacuum-based isothermal process. Characterized by scanning electron microscopy (SEM), the grain size of a CH₃NH₃PbI₃ film is on average 300 nm, while the largest grain size reaches over 800 nm, as shown in Fig. 1a. Surface roughness of the CH₃NH₃PbI₃ film was measured by atomic force microscopy (AFM). A smooth CH₃NH₃PbI₃ was also obtained by this process with a surface roughness less than RMS 17.4 nm, as shown in Fig. 1b, significantly smaller than the previous publication result.¹⁸ Large grain sizes would decrease photo-generated carrier scattering during transport inside the material. Low surface roughness would decrease carrier surface recombination when transporting between the interface of the electron/hole transport layer and the perovskite film.

Fig. 1 CH₃NH₃PbI₃ film on a FTO/c-TiO₂/m-TiO₂ substrate prepared by HPCVD method in a quartz tube at 100 °C for 3 h without thermal annealing: (a) a top-view SEM image with a grain size of about 300 nm; (b) tapping-mode AFM height image of the perovskite film with surface roughness of about RMS 17.4 nm; (c) corresponding 3D AFM image.

Quality of the solid PbI₂ precursor films also determines the resulting CH₃NH₃PbI₃ films. PbI₂ films were prepared by solution-based spin-coating process, and film thickness and uniformity were affected by rotating speed and vapor exhaustion condition, which would affect the recrystallization speed of PbI₂ material. As shown in Fig. 2b, the uniformity of PbI₂ film is poor when spin coating at 3000 rpm. By increasing rotating speed to 5000 rpm, the uniformity is greatly improved because of the faster crystallization of PbI₂, as shown in Fig. 2c. Higher spin-coating speed leads to acceleration of the PbI₂ precursor film crystallization, while the slower PbI₂ crystallization process results in a tree fork-shaped film. Our hypothesis is that uniform PbI₂ films would be obtained at the condition that the solvent in PbI₂ solution exhausts quickly and the PbI₂ crystallization process proceeds quickly. As shown in Fig. 2a, when nitrogen flow is applied during the spin-coating process, instead of increasing the spin-coating speed to increase PbI₂ recrystallization speed, PbI₂ film uniformity is improved under 3000 rpm (Fig. 2d) and further improved under 5000 rpm with the thickness of about 150–200 nm (Fig. 2e).

Fig. 2 (a) Schematic of optimized PbI₂ precursor film procedures and configuration of the subsequent HPCVD perovskite synthesis. Optimization of PbI₂ precursor films with spin-coating process: top view SEM images of PbI₂ films spun coated at (b) 3000 rpm and (c) 5000 rpm in a spin coater with the cover closed; top view SEM images of PbI₂ films spun coated at (d) 3000 rpm and (e) 5000 rpm with the cover open and nitrogen flow added during the rotation of the substrates. The pressure of the nitrogen flow was 0.5 bar.

Based on the optimized HPCVD method, high-performance CH₃NH₃PbI₃ solar cells were fabricated. Detailed information of the fabrication process is described in ESI.† As shown in Fig. 3a, cross-section of the perovskite solar cell was characterized by SEM. Uniform, 200 nm-thick CH₃NH₃PbI₃ film was laminated between the spiro-OMeTAD (organic hole transport material) layer and the TiO₂ (electron transport material) layer. A cross-sectional SEM image of the whole perovskite solar cell with vision field longer than 10 μm is shown in Fig. 3b, demonstrating the effectiveness of the HPCVD method in synthesizing perovskite films. X-ray diffraction (XRD) spectrum of the CH₃NH₃PbI₃ film is shown in Fig. 3c. Strong peaks in the XRD chart at 14.18°, 28.52°, and 31.96°, which correspond to (110), (220), and (310) Miller indices of CH₃NH₃PbI₃ perovskite crystal, respectively, indicate the tetragonal crystal structure of perovskite. The absence of the characteristic PbI₂ peak at 12.65° demonstrates the full reaction between the two precursors in the HPCVD process. From the absorbance spectrum of the CH₃NH₃PbI₃ film, as shown in Fig. 3d, bandgap of the perovskite material can be derived as 1.54 eV, in agreement with previous publications.^38,39

Fig. 3 (a) Cross-sectional SEM images of completed solar cells constructed from a HPCVD CH₃NH₃PbI₃ film with a thickness about 200 nm. (b) Cross-sectional SEM images under lower magnification of completed solar cells constructed from a HPCVD perovskite film. (c) XRD spectrum of the film after post-thermal annealing and (d) absorbance spectrum of the film with derived bandgap of 1.54 eV.

As shown in Fig. 4a, the best device synthesized by HPCVD displays great performance, with reverse scan: J_SC = 20.00 mA cm⁻², V_OC = 0.95 V, fill factor (FF) = 75%, PCE = 14.2%; forward scan: J_SC = 20.28 mA cm⁻², V_OC = 0.85 V, fill factor (FF) = 62%, PCE = 10.6%. In general, 27 devices exhibit reversing scan PCE from 6.2% to 14.2% with a mean value of 11.1%, as shown in Fig. 4b. External quantum efficiency (EQE) spectrum of the solar cell is shown in Fig. 4c. By integrating the EQE spectrum value with the AM 1.5G solar photon flux between 300 nm and 900 nm, short circuit current density of 16.5 mA cm⁻² can be derived, which is lower than J_SC from the J–V scanning and is probably due to the interface defects of the TiO₂ layer reported in previous publications.¹⁸

Fig. 4 (a) Current density–voltage curves of HPCVD CH₃NH₃PbI₃ solar cell measured under simulated AM 1.5 sunlight of 100 mW cm⁻² irradiance. (b) External quantum efficiency (EQE) spectrum of this device with integrated current density of 16.5 mA cm⁻². (c) Statistics of the efficiencies of 27 CH₃NH₃PbI₃ solar cells fabricated by the HPCVD method. Efficiencies of 17 devices are over 10%. (d) Performance of ten HPCVD solar cells was characterized after treatment at 80 °C for 50 h and 96 h. Average PCE decreased from 8.9% to 7.9% and 6.9% after 50 h and 96 h heat treatment, respectively. Detailed information is shown in Table 1 and S1.†

Finally, long-term stability of solar cells fabricated by HPCVD was characterized in a controlled harsh environment (facilities shown in Fig. S1†) by heating ten solar cells at 80 °C for 50 hours and 96 hours, continuously. As shown in Fig. 4d, the average PCE of the ten solar cells decayed gradually when the duration of the treatment increased, which is shown in Table 1. Detailed information of these solar cells is shown in Table S1.† The average PCE of the ten solar cells decreased from 8.9% to 7.9% and 6.9% after 50 h and 96 h treatment, still maintaining 78% of their initial efficiencies. Meanwhile, among the ten solar cells, PCEs of two cells remained stable after 96 h of harsh treatment, as shown in Table S1.† Perovskite solar cells whose organic hole transport material was poly(3-hexylthiophene) (P3HT) mixed with carbon nanotubes and polymethyl methacrylate (PMMA) have been reported to show stability at 80 °C for 96 hours with about 10% PCE decay; however, solar cells using Li-TSFI-doped spiro-OMeTAD as hole transport material in the same reference failed after the same heat treatment.⁴⁰ Still, the HPCVD perovskite solar cells demonstrated relative high stability compared with previously reported perovskite solar cells, the efficiency of which decreased rapidly when stored in nitrogen atmosphere.¹⁷

Table 1 Performance parameters of ten HPCVD solar cells before and after heat treatment at 80 °C for 50 h and 96 h^a

	Avg. JSC (mA cm^−2)	Avg. VOC (V)	Avg. FF	Avg. PCE (%)
a Detailed information is shown in Table S1.
Before heat treatment	16.55	0.96	0.54	8.9
After heated at 80 °C for 50 h	16.06	0.94	0.51	7.9
After heated at 80 °C for 96 h	14.53	0.92	0.48	6.9

---

Conclusions

In summary, high-quality perovskite (CH₃NH₃PbI₃) films were synthesized through a novel hybrid physical-chemical vapor deposition method for high-performance solar cells. Due to the vacuum and isothermal environment, high-quality and uniform CH₃NH₃PbI₃ films were obtained with grain sizes up to 800 nm and surface roughness around RMS 17.4 nm. Combined with an optimized spin-coating process for PbI₂ precursor film, a high-performance CH₃NH₃PbI₃ solar cell achieved high power conversion efficiency of up to 14.2%. After treatment in a controlled harsh environment (80 °C for 96 hours), solar cells maintained decently high efficiency, demonstrating that the HPCVD method was effective and stable. In our further research, more parameters of this HPCVD method and thermal annealing treatment will be modified to fabricate solar cells with better efficiency and stability. In future, the HPCVD perovskite solar cell could have great potential to be mass produced at low cost for large-scale applications.

Acknowledgements

This work was financially supported by the International Science and Technology Cooperation Program of Ministry of Science and Technology of the People's Republic of China (Grant No. 2013DFA51800).

Notes and references

1. M. A. Green, A. Ho-Baillie and H. J. Snaith, Nat. Photonics, 2014, 8, 506–514.
2. P. Gao, M. Gratzel and M. K. Nazeeruddin, Energy Environ. Sci., 2014, 7, 2448–2463.
3. A. Kojima, K. Teshima, Y. Shirai and T. Miyasaka, J. Am. Chem. Soc., 2009, 131, 6050–6051.
4. W. S. Yang, J. H. Noh, N. J. Jeon, Y. C. Kim, S. Ryu, J. Seo and S. I. Seok, Science, 2015, 348, 1234–1237.
5. J. Yang, F. Luo, T. S. Kao, X. Li, G. W. Ho, J. Teng, X. Luo and M. Hong, Light: Sci. Appl., 2014, 3, e185.
6. C. F. Guo, T. Sun, F. Cao, Q. Liu and Z. Ren, Light: Sci. Appl., 2014, 3, e161.
7. M. Memarian and G. V. Eleftheriades, Light: Sci. Appl., 2013, 2, e114.
8. H. S. Kim, C. R. Lee, J. H. Im, K. B. Lee, T. Moehl, A. Marchioro, S. J. Moon, R. Humphry-Baker, J. H. Yum, J. E. Moser, M. Gratzel and N. G. Park, Sci. Rep., 2012, 2, 591.
9. J. Burschka, N. Pellet, S. J. Moon, R. Humphry-Baker, P. Gao, M. K. Nazeeruddin and M. Gratzel, Nature, 2013, 499, 316–319.
10. N. J. Jeon, J. H. Noh, W. S. Yang, Y. C. Kim, S. Ryu, J. Seo and S. I. Seok, Nature, 2015, 517, 476–480.
11. N. J. Jeon, J. H. Noh, Y. C. Kim, W. S. Yang, S. Ryu and S. I. Seok, Nat. Mater., 2014, 13, 897–903.
12. P. Umari, E. Mosconi and F. de Angelis, Sci. Rep., 2014, 4, 4467.
13. G. Xing, N. Mathews, S. Sun, S. S. Lim, Y. M. Lam, M. Gratzel, S. Mhaisalkar and T. C. Sum, Science, 2013, 342, 344–347.
14. S. D. Stranks, G. E. Eperon, G. Grancini, C. Menelaou, M. J. Alcocer, T. Leijtens, L. M. Herz, A. Petrozza and H. J. Snaith, Science, 2013, 342, 341–344.
15. D. Shi, V. Adinolfi, R. Comin, M. Yuan, E. Alarousu, A. Buin, Y. Chen, S. Hoogland, A. Rothenberger, K. Katsiev, Y. Losovyj, X. Zhang, P. A. Dowben, O. F. Mohammed, E. H. Sargent and O. M. Bakr, Science, 2015, 347, 519–522.
16. M. Liu, M. B. Johnston and H. J. Snaith, Nature, 2013, 501, 395–398.
17. H. Zhou, Q. Chen, G. Li, S. Luo, T. B. Song, H. S. Duan, Z. Hong, J. You, Y. Liu and Y. Yang, Science, 2014, 345, 542–546.
18. Q. Chen, H. Zhou, Z. Hong, S. Luo, H. S. Duan, H. H. Wang, Y. Liu, G. Li and Y. Yang, J. Am. Chem. Soc., 2014, 136, 622–625.
19. M. Saba, M. Cadelano, D. Marongiu, F. Chen, V. Sarritzu, N. Sestu, C. Figus, M. Aresti, R. Piras, A. G. Lehmann, C. Cannas, A. Musinu, F. Quochi, A. Mura and G. Bongiovanni, Nat. Commun., 2014, 5, 5049.
20. L. Wang, C. McCleese, A. Kovalsky, Y. Zhao and C. Burda, J. Am. Chem. Soc., 2014, 136, 12205–12208.
21. V. W. Bergmann, S. A. Weber, F. Javier Ramos, M. K. Nazeeruddin, M. Gratzel, D. Li, A. L. Domanski, I. Lieberwirth, S. Ahmad and R. Berger, Nat. Commun., 2014, 5, 5001.
22. E. Edri, S. Kirmayer, S. Mukhopadhyay, K. Gartsman, G. Hodes and D. Cahen, Nat. Commun., 2014, 5, 3461.
23. J. Shi, J. Dong, S. Lv, Y. Xu, L. Zhu, J. Xiao, X. Xu, H. Wu, D. Li, Y. Luo and Q. Meng, Appl. Phys. Lett., 2014, 104, 063901.
24. J. Wei, Y. Zhao, H. Li, G. Li, J. Pan, D. Xu, Q. Zhao and D. Yu, J. Phys. Chem. Lett., 2014, 5, 3937–3945.
25. C. Wehrenfennig, M. Liu, H. J. Snaith, M. B. Johnston and L. M. Herz, Energy Environ. Sci., 2014, 7, 2269.
26. C. Wehrenfennig, G. E. Eperon, M. B. Johnston, H. J. Snaith and L. M. Herz, Adv. Mater., 2014, 26, 1584–1589.
27. F. Huang, Y. Dkhissi, W. Huang, M. Xiao, I. Benesperi, S. Rubanov, Y. Zhu, X. Lin, L. Jiang, Y. Zhou, A. Gray-Weale, J. Etheridge, C. R. McNeill, R. A. Caruso, U. Bach, L. Spiccia and Y.-B. Cheng, Nano Energy, 2014, 10, 10–18.
28. Y. Dkhissi, F. Huang, S. Rubanov, M. Xiao, U. Bach, L. Spiccia, R. A. Caruso and Y.-B. Cheng, J. Power Sources, 2015, 278, 325–331.
29. J. You, Y. Yang, Z. Hong, T.-B. Song, L. Meng, Y. Liu, C. Jiang, H. Zhou, W.-H. Chang, G. Li and Y. Yang, Appl. Phys. Lett., 2014, 105, 183902.
30. B.-S. Kim, T.-M. Kim, M.-S. Choi, H.-S. Shim and J.-J. Kim, Org. Electron., 2015, 17, 102–106.
31. S. T. Ha, X. F. Liu, Q. Zhang, D. Giovanni, T. C. Sum and Q. H. Xiong, Adv. Opt. Mater., 2014, 2, 838–844.
32. D. J. Lewis and P. O'Brien, Chem. Commun., 2014, 50, 6319–6321.
33. M. R. Leyden, L. K. Ono, S. R. Raga, Y. Kato, S. H. Wang and Y. B. Qi, J. Mater. Chem. A, 2014, 2, 18742–18745.
34. P. Luo, Z. Liu, W. Xia, C. Yuan, J. Cheng and Y. Lu, ACS Appl. Mater. Interfaces, 2015, 7, 2708–2714.
35. W.-J. Yin, J.-H. Yang, J. Kang, Y. Yan and S.-H. Wei, J. Mater. Chem. A, 2015, 3, 8926–8942.
36. Y. Peng, G. Jing and T. Cui, J. Mater. Chem. A, 2015, 3, 12436–12442.
37. Y. Peng, G. Jing and T. Cui, in MRS Proceedings, Cambridge Univ Press, 2015, pp. mrss15–2135352.
38. G. Niu, X. Guo and L. Wang, J. Mater. Chem. A, 2015, 3, 8970–8980.
39. W. Nie, H. Tsai, R. Asadpour, J. C. Blancon, A. J. Neukirch, G. Gupta, J. J. Crochet, M. Chhowalla, S. Tretiak, M. A. Alam, H. L. Wang and A. D. Mohite, Science, 2015, 347, 522–525.
40. S. N. Habisreutinger, T. Leijtens, G. E. Eperon, S. D. Stranks, R. J. Nicholas and H. J. Snaith, Nano Lett., 2014, 14, 5561–5568.

---

Footnotes

† Electronic supplementary information (ESI) available: Additional experimental introduction and data. See DOI: 10.1039/c5ra19343e

‡ The first two authors contributed equally to this work.

---