Highly reproducible perovskite solar cells with excellent CH₃NH₃PbI_3−xCl_x film morphology fabricated via high precursor concentration

Abstract

To meet the large-scale fabrication technology of perovskite solar cells in the future, a novel method is proposed herein to achieve an excellent and highly reproducible CH₃NH₃PbI_3−xCl_x film based on a high concentration spinning process. Pinholes in the CH₃NH₃PbI_3−xCl_x film were substantially reduced, leading to better uniformity and film coverage. The precursor ions and molecules promoted nucleation and growth while releasing by-products gradually, which was beneficial to continuous growth and crystallization of the perovskite film in high concentration conditions and then resulted in significantly enhanced optoelectronic characteristics and device performance. Up to 13.8% of power conversion efficiency had been achieved by employing the FTO/bl-TiO₂/mp-TiO₂/CH₃NH₃PbI_3−xCl_x/Spiro-MeOTAD/Au device structure under AM1.5 global spectral irradiation (100 mW cm⁻²). This method offered an efficacious and simple strategy for a highly reproducible large-area fabrication process of low-cost perovskite solar cells.

---

Introduction

Organic/inorganic hybrid perovskite structure materials possess many advantageous properties such as long exciton diffusion lengths, high molar absorption coefficient, low cost and so on. They exhibit outstanding light capture capacity and electrical conductivity.^1–3 A new family of solution-processable perovskite structure materials based on organolead trihalide compounds (CH₃NH₃PbX₃, X = Cl, Br, I) have become very attractive as they are promising for cost competitive solar cells which are currently under rapid development as compared with traditional thin film photovoltaic devices.^4,5 In recent years, lead halide perovskite thin film photovoltaic devices have shown the emergence process with power conversion efficiency (PCE) of over 20%.⁶ Besides their high conversion efficiency, in order to meet the commercial demand for perovskite solar cell modules, a challenge to solution-processed perovskite thin films is realizing highly reproducible perovskite solar cells with perfect thin film morphology and interfaces, which can be of great benefit to the large-area fabrication process of low-cost perovskite solar cells.^4,7,8

For typical solution deposition method to prepare morphology-controlled perovskite film, great efforts have been made to improve the film quality via reducing the perovskite film pin-hole defect density and enhancing compactness. Such morphology-controlled perovskite film deposition processes have been carried out in interface modification,^9–13 annealing engineering,^3,14–19 additive-enhanced methods^20–28 and solvent engineering^29–33 etc. Nevertheless, these methods render it challenging to achieve high efficiency as well as high reproducibility due to the unfavorable effects during preparation and annealing processes, and are not suitable for the large-scale fabrication technology of perovskite solar cells in the future. NG Park's group achieved highly reproducible perovskite solar cells by employing Lewis base adduct of PbI₂ and diethyl ether was used to remove solvent DMF.³⁴ Han research group overcame the problem of incomplete conversion and uncontrolled particle size of perovskite in the absence of mesoporous scaffolds by retarding the crystallization of PbI₂.³⁵ Bert Conings reported efficient hybrid solar cells based on the mixed metal halide perovskite absorber CH₃NH₃PbI₂Cl with device PCE exceeded 10% and high reproducibility resulted from a thin film sandwich approach.³⁶ However, generally in previous cases, such highly reproducibly controlled perovskite films require relatively complicate and precise condition control during processes, such as the solvent washing process.³⁴ Therefore, realizing an effective and simple method for preparing reproducible film with well controlled morphology is highly imperative.

In this work, a simple strategy to improve the perovskite film morphology quality was proposed. We prepared the CH₃NH₃PbI_3−xCl_x film via increasing precursor concentration, which could effectively increase the film morphology coverage and crystallinity. The structural and optical properties of the films prepared based on 50 wt% precursor concentration were characterized by the scanning electron microscope (SEM), X-ray diffraction (XRD) and photoluminescence spectra in comparison with the films prepared from 40 wt% and 45 wt% precursor concentration solution. Results showed that this method can effectively reduce film pinhole proportion, leading to better morphology evolution. Then the photovoltaic devices were prepared. Their photovoltaic performance parameters were discussed in various concentration conditions. The photovoltaic devices based 50 wt% precursor concentration showed the average PCE of 13.2% under AM1.5 global spectral irradiation (100 mW cm⁻²).

Experimental

Device fabrication

The solar cell structure was FTO/bl-TiO₂/mp-TiO₂/CH₃NH₃PbI_3−xCl_x/Spiro-MeOTAD/Au. The FTO (sheet resistance 10 Ω square⁻¹) substrate was cleaned with soap, deionized water, acetone and ethanol sequentially. The methods of preparing compact TiO₂ (bl-TiO₂) and mesoporous TiO₂ (mp-TiO₂) layers had been described in detail in our previous report.¹⁵ The perovskite films were prepared by one-step spin coating procedure via the hybrid solution. Typically, the CH₃NH₃I (prepared by the reported procedure³⁷) was mixed with PbCl₂ (Alfar Aesar) and PbI₂ (Alfar Aesar). The mole ratio of PbCl₂ : PbI₂ : CH₃NH₃I was controlled to be 1 : 1 : 4 in DMF (Alfar Aesar). The hybrid solution was kept being stirred at room temperature for 2 h to obtain perovskite precursor solution with varied concentration of 40 wt%, 45 wt% and 50 wt%, respectively. And lastly, the precursor solution was filtered with a PTFE filter (0.45 μm pore size) to obtain more transparent perovskite precursor solution.

To deposit perovskite film, TiO₂ substrates and precursor solution were transferred to the glove box with water content less than 50 ppm. Then the TiO₂ substrates and precursor solution were heated at 70 °C for 1 h. The perovskite precursor solution was spin-coated on the mp-TiO₂ substrate with spinning speed of 4000 rpm for 60 s. The coated perovskite films were then placed in drying cabinet at room temperature for 1 h (under relative humidity of 20%), followed by thermal annealing on the hot plate at 100 °C for 75 min under relative humidity below 5%. The color of perovskite film was brown black in the end of preparation.

For depositing hole transport layer, Spiro-MeOTAD (Luminescence Corp) was mixed in chlorobenzene solution (73 mg/1 mL), including 20 μL of tert-butylpyridine (t-BP) and 17.5 μL of Li-bis(trifluoromethanesulfonyl)-imide (Li-TFSI) dissolved in acetonitrile (520 mg mL⁻¹). The Spiro-MeOTAD was spin-coated on perovskite film at 2000 rpm for 30 s. And then a layer of 100 nm Au was deposited on the top of the devices by thermal evaporation.

Characterization

Transmittance spectra of perovskite films were recorded by a Perkin Elmer Lambda 950 UV-Vis Spectrometer. XRD was performed by DX-2600 X-ray diffractometer (Dandong Fangyuan Instrument Company) with Cu-Kα radiation (scan range from 10° to 70°). Film morphology was observed by field emission scanning electron microscope (Hitachi S-4300). Time-resolved photoluminescence of perovskite films was characterized by an Edinburgh Instruments FLS 980 fluorescence spectrometer. The I–V characteristic was measured with Keithley 2400 SourceMeter under simulated AM1.5 G (Sun 2000 solar simulator, ABET technologies) with irradiation power density of 100 mW cm⁻² and calibrated with a GaAs reference cell. Device area was typically 0.07 cm² and the area of the devices for EQE measurements (QEX10, PV Measurements, Inc.) was 0.18 cm².

Results and discussion

Surface morphological features of perovskite films prepared with different precursor solution concentration were observed by SEM, as shown in Fig. 1. On the other hand, the thickness of the corresponding perovskite layers is listed in ESI Table S1.† As can be seen from the results, the surface morphology of solution-processed perovskite films is strongly dependent on precursor solution concentration. Pin-hole defects often appeared in perovskite thin films prepared by conventional 40 wt% concentration, which led to decreased film compactness and undesirable photovoltaic performance as well. Our previous studies showed that the perovskite films quality were significantly improved by regulating the mole ratio of precursor solution and controlling grain growth process by two-step annealing method.¹⁵ However, this method make it challenging for low cost large-scale preparation processes in the future. To overcome this problem, we deposited perovskite films through a simple strategy to realize excellent CH₃NH₃PbI_3−xCl_x films morphology. Fig. 1(a) and (d) show the perovskite film with different scale bars prepared with 40 wt% precursor solution concentration. Relatively small grains with average grain size of 560 nm were observed. Pinholes observed in the perovskite films are due to byproducts (e.g. CH₃NH₃, HCl vapor) are releasing through chemistry reaction during film annealing afterward that resulted in some grain boundary cracks. In the following case, the average grain size is about 670 nm when the perovskite thin film was made from solution with concentration of 45 wt%, which is corresponding to SEM images in Fig. 1(g). Similarly, these pinholes were also observed in Fig. 1(b and e). It should be noted that the average pinhole size decreased from 135 nm to 110 nm as the precursor concentration increased from 40 wt% to 45 wt% as shown in Fig. 1(h). Meanwhile, the percentage of large-size pinholes beyond 150 nm decreased from 38.7% to 17.2%. The results indicated that elevating perovskite precursor solution concentration can substantially reduce film pinhole proportion, resulting in better morphology.³⁶ To further improve the morphology quality, the perovskite film is prepared by precursor solution with concentration of 50 wt%. The grain size increased to about 900 nm as shown in Fig. 1(g). Nearly continuous and compact perovskite film was observed based on high-concentration precursor solution method in Fig. 1(c and f). In the film growth and crystallization process, nucleation started from precursor ions and molecules and they grew gradually while releasing byproducts, which could promote continuous growth and crystallization of perovskite film in high concentration solution. The uniform, pinhole-free and highly crystallized thicker perovskite film was achieved on mp-TiO₂ substrate. Additionally, we tried to increase the precursor solution concentration further to check whether higher concentration could improve the film quality more or not. It was found that when the precursor solution concentration was elevated to 55 wt% the film morphology showed only slightly difference between that prepared from 50 wt% concentration solution, as shown in ESI Fig. S2.† In this case, the thickness of the film became thicker, which has an influence on the performance of the device.³⁶ Moreover, the as-prepared perovskite films showed mirror-like reflection behavior as shown in inset Fig. 1(a–c), which indicated that the films are more stable for efficient solar cells.²⁹

Fig. 1 (a–f) SEM images of the perovskite films prepared from varied precursor solution concentration: (a and d) 40 wt%, (b and e) 45 wt% and (c and f) 50 wt%, respectively. Insert: optical images of the films. (g) Grain size distribution corresponding to SEM images. (h) Pin-hole size distribution in the perovskite films.

XRD were employed to study the film crystallinity. Fig. 2 is the XRD spectra of CH₃NH₃PbI_3−xCl_x films on the mp-TiO₂ substrates fabricated from different solution concentration. The diffraction peaks at 14.2°, 28.5° and 32° are assigned to the (110), (220) and (310) peaks, respectively, which contributes to the tetragonal crystal structure.³⁸ In comparison with the films prepared from 40 wt% and 45 wt% solution, the film prepared from 50 wt% solution showed stronger diffraction peaks with highly preferred growth orientation along (110) direction. This exhibited that the higher precursor solution concentration induced larger grain growth and enhance crystallization process, resulting in improved crystallinity of perovskite film.

Fig. 2 XRD patterns of the perovskite films prepared from solution with 40 wt%, 45 wt% and 50 wt% concentration, respectively.

The transmittance spectra and absorption coefficient of the perovskite films are shown in Fig. 3. All films showed high absorption over the visible region with band-edge absorption onset at about 800 nm, as revealed by Fig. 3(a). The transmittance of the film prepared from 50 wt% concentration solution was lower than that from 40 wt% and 45 wt% solution, which could be ascribed to better perovskite film crystallinity and surface coverage. The optical band gaps (E_g) could be obtained from the following Tauc equation by assuming a direct transition:

(αhν)² = A(hν − E_g), (1)

where A, α and h, are a constant, absorption coefficient and Planck constant, respectively. E_g was estimated to be 1.58 eV which is consistent with reported results.³⁹ The absorption coefficient of the film deposited from solution with 50 wt% concentration was greatly elevated as illustrated in Fig. 3(b), which should be due to the enhanced crystallinity and surface coverage, leading to strong light harvesting capacity.

Fig. 3 (a) Transmittance spectra and (b) absorption coefficient of the perovskite films fabricated from solution with 40 wt%, 45 wt% and 50 wt% concentration, respectively.

Additionally, to further investigate the recombination processes of CH₃NH₃PbI_3−xCl_x films. The photoluminescence spectra and time-resolved photoluminescence spectra were measured based on time correlated single photon counting as shown in Fig. 4. The wavelength of maximum emission of all the CH₃NH₃PbI_3−xCl_x films at 774 nm can be observed in Fig. 4(a), which is consistent with our previous reports.¹⁵ It can be assigned to direct band-edge transition from the valence band to the conduction band, which is corresponding to the optical band gap calculated by Tauc plots with the value of ∼1.58 eV. Fig. 4(b) is the time-resolved photoluminescence spectra of all samples. Similarly, the lifetimes were fitted with three exponential decay curves (the longer lifetime was adopted for comparison).⁴⁰ The PL decay long carrier lifetime of the film deposited from 50 wt% precursor concentration solution had long lifetime of 591 ns, which was longer than that of the films prepared from low-concentration solution (265 ns for 40 wt% and 466 ns for 45 wt% solution). It is comparable to the lifetime for CH₃NH₃PbI_3−xCl_x film in previous reports.^11,39 Long carrier lifetime is beneficial to promote longer carrier diffusion lengths for thicker perovskite film, which was attributed to the reduced defect density and then suppressed non-radiative recombination channels.⁴⁰

Fig. 4 (a) Photoluminescence (b) time-resolved photoluminescence at 774 nm of perovskite films prepared from varied precursor solution concentration of 40 wt%, 45 wt% and 50 wt% respectively.

The dependence of solar cell performance on film morphology was investigated. With device structure of FTO/bl-TiO₂/mp-TiO₂/CH₃NH₃PbI_3−xCl_x/Spiro-MeOTAD/Au. The cross-section SEM images of the devices corresponding to varied precursor solution concentration were shown in Fig. S3.† Table 1 shows the photovoltaic parameters of the devices with perovskite layers made from solution of varied precursor concentration. Fig. 5 exhibits typical I–V characteristics measured under AM1.5 G, 100 mW cm⁻² light illumination, external quantum efficiency (EQE) and integrated photocurrent density of the devices were derived from the I–V characteristics. Similarly, the device appeared hysteresis behavior during the current–voltage measurement, depending on the scan rate and scan directions. The devices were recorded with delay time 50 ms (stepwise scanning of 10 mV) from 1 V to −0.1 V under reverse and forward scanning conditions. Specifically, the devices with CH₃NH₃PbI_3−xCl_x perovskite layers fabricated with 50 wt% precursor concentration solution exhibited average short-circuit current (J_sc) of 19.5 mA cm⁻², open circuit voltage (V_oc) of 0.92 V, fill factor (FF) of 73.5%, corresponding to PCE of 13.2% under reverse scan direction and J_sc of 20.1 mA cm⁻², V_oc of 0.87 V, FF of 70.1% and PCE of 12.3% under forward scan direction. The performance is significantly better than that of 40 wt% (J_sc of 15.46 mA cm⁻², V_oc of 0.84 V and FF of 65.1% under reverse scan direction, J_sc of 16.01 mA cm⁻², V_oc of 0.79 V, FF of 60.1% and PCE of 7.67% under forward scan direction) and 45 wt% (J_sc of 18 mA cm⁻², V_oc of 0.88 V and FF of 71.8% under reverse scan direction, J_sc of 17.97 mA cm⁻², V_oc of 0.84 V, FF of 67% and PCE of 10.1% under forward scan direction) precursor concentration samples, as shown in Fig. 5(a). The average PCE obtained for the 50 wt% samples are quite higher than others, which can be derived from better perovskite film morphology coverage and crystallinity. The properties of surface coverage, crystallinity and grain size of the perovskite films play an important role in photovoltaic performance. According to the SEM images of different perovskite films in Fig. 1, it is obvious that pinhole size and density was significantly dependent on precursor concentration. Poor surface coverage lowers the film and device preparation processes reproducibility, leading to undesirable photovoltaic performance. The photocurrent density and PCE exhibited small standard deviation as summarized in Table 1. It is reasonable that the reproducibility of device performance can be improved by the controlled CH₃NH₃PbI_3−xCl_x film morphology. Pinholes were almost eliminated in the film deposited from 50 wt% precursor concentration solution. Such pinholes are unfavorable for photo-induced carrier generation and charge transport along the vertical direction,⁴¹ leading to lower J_sc and PCE. Fig. S4† shows the histograms of device efficiency for 40 cells in 50 wt% precursor solution condition. It indicates that increasing precursor concentration can effectively improve the reproducibility of high efficiency perovskite solar cells.

Table 1 The photovoltaic performance of perovskite solar cells fabricated from varied precursor concentration solution. (Based on 20 cells of each group.)

Concentration	Jsc (mA cm^−2)	Voc (V)	FF (%)	PCE (%)
40 wt%	15.46 ± 2.1	0.84 ± 0.4	65.1 ± 5.3	8.4 ± 2.2
45 wt%	18 ± 0.9	0.88 ± 0.5	71.8 ± 4.8	11.3 ± 0.8
50 wt%	19.5 ± 0.8	0.92 ± 0.4	73.5 ± 6.0	13.2 ± 0.6

---

Fig. 5 (a) Typical J–V curve of the devices measured under AM1.5 simulated sun light for reverse and forward scan; (b) EQE spectra and integrated photocurrent density.

High-concentration spin method is beneficial to enhance the film surface coverage, crosslinking of grains and crystallinity. This suggested that the perovskite layers showed excellent light harvesting capacity due to higher surface coverage, contributing to higher photo-generated current. Fig. 5(b) is the EQE response and integrated photocurrent density of the devices. As can be seen from this figure, all the devices' spectral response covered the visible region with band-edge onset around 800 nm. Moreover, the TiO₂/CH₃NH₃PbI_3−xCl_x/Spiro-MeOTAD interfaces and layers are highly efficient to extract charges from 350 to 800 nm. For the devices prepared with high precursor concentration (50 wt%), they had slightly higher photocurrent in the range of 500–800 nm than others due to increased absorption and crystallinity of the perovskite films. The perovskite films produced from low precursor concentration solution (40 wt% and 45 wt%) can easily cause larger pinholes with an incomplete surface coverage, then hinder the carrier transport in the films and also increase the charge recombination channels which led to decreased photo-generated current and lowered EQE spectra.

Additionally, Fig. S5† is the J–V curve and EQE spectra of the device fabricated with 55 wt% concentration measured under AM1.5 simulated sun light. The device exhibits J_sc of 20.46 mA cm⁻², V_oc of 0.84 V, FF of 68.7% and PCE of 11.9%. This proved that further elevating concentration resulted in higher J_sc because of thicker absorption layer of perovskite film. However, too high concentration precursor solution causes thicker perovskite film, which may also increase the carrier recombination and internal resistance of the device and lead to lower V_oc and FF.

Fig. 6(a–d) are the box charts of detailed photovoltaic performance comparison. The device parameters showed strong dependence on precursor solution concentration of perovskite films. As mentioned above, the film morphology quality was improved with increasing concentration, leading to higher photo-generated current. Poor surface coverage induced large pinholes formation and brought about contact between the Spiro-OMeTAD layer and TiO₂ substrate.⁴¹ The shunt current paths could not be eliminated. Higher surface coverage and grain size could reduce the defect density at grain boundaries and decrease the carrier recombination channels. The V_oc and FF was improved significantly with increased precursor solution concentration, as shown in Fig. 6(b).

Fig. 6 Box charts of photovoltaic performances of solar cells prepared from varied precursor concentration solution ((a) current density, (b) V_oc, (c) FF and (d) PCE, respectively).

The diode equivalent circuit model equation based on heterojunction solar cell was applied to analyze the J–V characteristics of the devices. For perovskite photovoltaic devices, I–V characteristics equation can be described by^42,43

where, J, J_sc, I₀, e, n, V, R_s, k, R_sh and T represent current density on the external load, photocurrent, reverse saturation current, electric charge, ideality factor, bias potential, series resistance, Boltzmann constant, shunt resistance and absolute temperature, respectively. For all the three series samples, the shunt resistance was around 2972 Ω cm².^42,44 After derivative operation, eqn (2) can also be written as⁴⁴

Based on eqn (3), R_s can be derived by fitting the curves of versus in Fig. 7(a). R_s and n can be estimated from the intercept and slope of fitted linear plots and were summarized in Table 2. The series resistance of the three series solar cells were calculated to be 1.36 Ω cm², 0.92 Ω cm² and 0.49 Ω cm² respectively as the precursor solution concentration for depositing perovskite films increased from 40 wt% to 45 wt% and 50 wt%. The ideality factor of corresponding devices was 3.20, 2.73 and 2.65, respectively. The reduced series resistance and diode ideality factor might be brought about by improved quality of perovskite film and their neighboring interfaces as the precursor solution concentration was elevated. Similarly, recombination current density can be derived from linear fitting of ln(J + J_sc − V/R_sh) versus V − R_sJ, as shown in Fig. 7(b). As the precursor solution concentration rose from 40 wt% to 50 wt% by intervals of 5 wt%, the recombination current density of the corresponding three series cells was lowered from 6.6 × 10⁻⁶ mA cm⁻² to 4.8 × 10⁻⁸ mA cm⁻² and 2.4 × 10⁻⁸ mA cm⁻², respectively. The smaller J₀ was a sign of substantially suppressed charge recombination loss process due to high-concentration precursor solution method which can improve crystallinity and grain size, contributing to higher V_oc and FF.

Fig. 7 (a) Plots of versus and the linear fitted curves. (b) Plots of ln(J + J_sc − V/R_sh) versus (V − R_sJ) and the linear fitted curves.

Table 2 R_s, n and J₀ derived from J–V characteristics

Concentration	Rs (Ω cm^2)	n	J0 (mA cm^2)
40 wt%	1.36	3.2	6.6 × 10^−6
45 wt%	0.92	2.73	4.8 × 10^−8
50 wt%	0.49	2.65	2.4 × 10^−8

---

Conclusion

In conclusion, we proposed and studied a simple and effective strategy to produce highly reproducible CH₃NH₃PbI_3−xCl_x film morphology by applying high concentration precursor solution. It could significantly improve the film morphology quality, crystallinity and interface quality that is comparable with that made from complicated processes. The XRD, UV-Vis and SEM characterizations indicated that the film had excellent uniformity, higher absorption coefficient and larger grain size compared with the reference films. Further, the pin-hole defect density of perovskite film was sharply decreased, which could reduce the non-radiative recombination paths. It contributed to large photovoltaic performance enhancement of the devices and 13.8% of PCE had been achieved with high reproducibility. This method offered an effective strategy for highly reproducible perovskite solar cells with excellent morphology, which might facilitate the large area fabrication process of low cost perovskite solar cells.

Acknowledgements

This work was financially supported by National High Technology Research and Development Program of China (No. 2011AA050515), and the Innovation Program of Sichuan University (No. 201510611374, No. 201510610799 and No. 201510610437).

References

1. H. S. Jung and N.-G. Park, Small, 2014, 11, 10–25.
2. M. A. Green, A. Ho-Baillie and H. J. Snaith, Nat. Photonics, 2014, 8, 506–514.
3. J. Liu, C. Gao, X. He, Q. Ye, L. Ouyang, D. Zhuang, C. Liao, J. Mei and W. Lau, ACS Appl. Mater. Interfaces, 2015, 7, 24008–24015.
4. H. J. Snaith, J. Phys. Chem. C, 2013, 4, 3623–3630.
5. M.-E. Ragoussi and T. Torres, Chem. Commun., 2015, 51, 3957–3972.
6. 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.
7. W. Chen, Y. Wu, Y. Yue, J. Liu, W. Zhang, X. Yang, H. Chen, E. Bi, I. Ashraful, M. Grätzel and L. Han, Science, 2015, 350, 944–948.
8. S. D. Stranks, P. K. Nayak, W. Zhang, T. Stergiopoulos and H. J. Snaith, Angew. Chem., Int. Ed., 2015, 54, 3240–3248.
9. Z.-K. Wang, M. Li, D.-X. Yuan, X.-B. Shi, H. Ma and L.-S. Liao, ACS Appl. Mater. Interfaces, 2015, 7, 9645–9651.
10. C. Sun, Q. Xue, Z. Hu, Z. Chen, F. Huang, H. L. Yip and Y. Cao, Small, 2015, 11, 3344–3350.
11. 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.
12. L. Zuo, Z. Gu, T. Ye, W. Fu, G. Wu, H. Li and H. Chen, J. Am. Chem. Soc., 2015, 137, 2674–2679.
13. H. Dong, X. Guo, W. Li and L. Wang, RSC Adv., 2014, 4, 60131–60134.
14. S. R. Raga, M.-C. Jung, M. V. Lee, M. R. Leyden, Y. Kato and Y. Qi, Chem. Mater., 2015, 27, 1597–1603.
15. D. Liu, L. Wu, C. Li, S. Ren, J. Zhang, W. Li and L. Feng, ACS Appl. Mater. Interfaces, 2015, 7, 16330–16337.
16. L. Huang, Z. Hu, J. Xu, K. Zhang, J. Zhang and Y. Zhu, Sol. Energy Mater. Sol. Cells, 2015, 141, 377–382.
17. Z. Xiao, Q. Dong, C. Bi, Y. Shao, Y. Yuan and J. Huang, Adv. Mater., 2014, 26, 6503–6509.
18. H. L. Hsu, C. P. Chen, J. Y. Chang, Y. Y. Yu and Y. K. Shen, Nanoscale, 2014, 6, 10281–10288.
19. Z. Ren, A. Ng, Q. Shen, H. C. Gokkaya, J. Wang, L. Yang, W.-K. Yiu, G. Bai, A. B. Djurisic, W. W.-f. Leung, J. Hao, W. K. Chan and C. Surya, Sci. Rep., 2014, 4, 6752–6757.
20. C.-C. Chueh, C.-Y. Liao, F. Zuo, S. T. Williams, P.-W. Liang and A. K. Y. Jen, J. Mater. Chem. A, 2015, 3, 9058–9062.
21. K. Sun, J. Chang, F. H. Isikgor, P. Li and J. Ouyang, Nanoscale, 2015, 7, 896–900.
22. X. Gong, M. Li, X.-B. Shi, H. Ma, Z.-K. Wang and L.-S. Liao, Adv. Funct. Mater., 2015, 25, 6671–6678.
23. C.-Y. Chang, C.-Y. Chu, Y.-C. Huang, C.-W. Huang, S.-Y. Chang, C.-A. Chen, C.-Y. Chao and W.-F. Su, ACS Appl. Mater. Interfaces, 2015, 7, 4955–4961.
24. P.-W. Liang, C.-Y. Liao, C.-C. Chueh, F. Zuo, S. T. Williams, X.-K. Xin, J. Lin and A. K. Y. Jen, Adv. Mater., 2014, 26, 3748–3754.
25. L. Yang, J. Wang and W. W.-F. Leung, ACS Appl. Mater. Interfaces, 2015, 5, 14614–14619.
26. Y. Xu, L. Zhu, J. Shi, S. Lv, X. Xu, J. Xiao, J. Dong, H. Wu, Y. Luo, D. Li and Q. Meng, ACS Appl. Mater. Interfaces, 2015, 7, 2242–2248.
27. Y. Zhao and K. Zhu, J. Phys. Chem. C, 2014, 118, 9412–9418.
28. Z. Zhang, X. Yue, D. Wei, M. Li, P. Fu, B. Xie, D. Song and Y. Li, RSC Adv., 2015, 5, 104606–104611.
29. W. Li, J. Fan, J. Li, Y. Mai and L. Wang, J. Am. Chem. Soc., 2015, 137, 10399–10405.
30. N. J. Jeon, J. H. Noh, Y. C. Kim, W. S. Yang, S. Ryu and S. I. Seok, Nat. Mater., 2014, 13, 897–903.
31. Y. Zhao and K. Zhu, J. Phys. Chem. C, 2014, 5, 4175–4186.
32. R. Kang, J.-E. Kim, J.-S. Yeo, S. Lee, Y.-J. Jeon and D.-Y. Kim, J. Phys. Chem. C, 2014, 118, 26513–26520.
33. M. Lv, X. Dong, X. Fang, B. Lin, S. Zhang, J. Ding and N. Yuan, RSC Adv., 2015, 5, 20521–20529.
34. N. Ahn, D.-Y. Son, I.-H. Jang, S. M. Kang, M. Choi and N.-G. Park, J. Am. Chem. Soc., 2015, 137, 8696–8699.
35. Y. Wu, A. Islam, X. Yang, C. Qin, J. Liu, K. Zhang, W. Peng and L. Han, Energy Environ. Sci., 2014, 7, 2934–2938.
36. B. Conings, L. Baeten, C. De Dobbelaere, J. D'Haen, J. Manca and H.-G. Boyen, Adv. Mater., 2014, 26, 2041–2046.
37. J. H. Im, I. H. Jang, N. Pellet, M. Gratzel and N. G. Park, Nat. Nanotechnol., 2014, 9, 927–932.
38. Y. Zhao and K. Zhu, Chem. Soc. Rev., 2016, 45, 655–689.
39. D. Wang, Z. Liu, Z. Zhou, H. Zhu, Y. Zhou, C. Huang, Z. Wang, H. Xu, Y. Jin, B. Fan, S. Pang and G. Cui, Chem. Mater., 2014, 26, 7145–7150.
40. J. You, Z. Hong, Y. Yang, Q. Chen, M. Cai, T.-B. Song, C.-C. Chen, S. Lu, Y. Liu, H. Zhou and Y. Yang, ACS Nano, 2014, 8, 1674–1680.
41. L. Huang, Z. Hu, G. Yue, J. Liu, X. Cui, J. Zhang and Y. Zhu, Phys. Chem. Chem. Phys., 2015, 17, 22015–22022.
42. 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.
43. 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.
44. J. Qing, H.-T. Chandran, Y.-H. Cheng, X.-K. Liu, H.-W. Li, S.-W. Tsang, M.-F. Lo and C.-S. Lee, ACS Appl. Mater. Interfaces, 2015, 7, 23110–23116.

---

Footnote

† Electronic supplementary information (ESI) available: The thickness PSC films, the SEM images of the perovskite films prepared with 55 wt% concentration, the cross-sectional SEM images of the devices, histograms of device efficiency and J–V curve and EQE spectra of the device fabricated with 55 wt% concentration. See DOI: 10.1039/c6ra07359j

---