Morphology fixing agent for [6,6]-phenyl C₆₁-butyric acid methyl ester (PC₆₀BM) in planar-type perovskite solar cells for enhanced stability

Abstract

Here, we report that the delamination of [6,6]-phenyl C₆₁-butyric acid methyl ester (PC₆₀BM) from the CH₃NH₃PbI₃ (MAPbI₃) layer is a critical reason for the degradation of inverted perovskite solar cells (PSCs). In this work, an ∼10 nm-thick titanium oxide (TiO_x) layer can function as a morphology-fixing agent for preventing PC₆₀BM delamination, which in turn induces greatly improved long-term stability of the PSCs. In addition, the devices with the TiO_x layer exhibited increased open circuit voltages, current densities, and fill factors, which correspond to improved initial device efficiency. Surface morphology changes of the PC₆₀BM layer on the MAPbI₃ layer during long-term operation of PSCs were confirmed by SEM and AFM, and were also found to be correlated with various electrical properties of the devices such as the hole mobility and charge generation and extraction efficiencies from the space-charge-limited current (SCLC) measurements and J_ph versus V_int characterizations.

---

Introduction

Organic–inorganic hybrid perovskites solar cells (PSCs) have attracted considerable interest for the last few years owing to their advantages of high performance and low-cost fabrication through the use of a simple solution process.¹ In particular, methylammonium lead halide (MAPbX₃, where MA is CH₃NH₃ and X is Cl, Br or I) perovskites have been expected to make rapid progress for use in solar cells by bringing about broad light absorption that is tunable by control of the chemical composition,² and due to their excellent photovoltaic properties such as long exciton diffusion lengths.^3–7 Planar structures of PSCs based on MAPbX₃ have been fabricated through the solution process and their power conversion efficiencies have attained over 15%, which came about through various efforts to improve crystallinity and gain large and uniform grain sizes in the perovskite film such as the solvent washing process, the annealing process, the doping process, etc.^8–14 Conventional PSCs, however, are composed of metal oxides (zinc oxides or titanium oxides as an electron transport layer) which require high-temperature processes based on the structure of dye-sensitized solar cells (e.g., sintering of mesoporous TiO₂ particles). But, since metal oxides aren't needed in the inverted PSCs as a blocking layer, the hybrid PSCs can become viable alternatives to conventional PSCs because of their low-temperature and comparatively very low-cost fabrication processes.

The [6,6]-phenyl C₆₁-butyric acid methyl ester (PC₆₀BM) is widely used as an electron transport layer (ETL) in inverted PSCs due to its outstanding charge transport properties. Dissolved polar-solvent-based PC₆₀BM solution can generate thin-film stacking on top of the perovskite layer from various solution processes. However, as the solvent of PC₆₀BM is gradually dried during thin-film fabrication, deficient adhesion between the perovskite and PC₆₀BM can cause a crevice due to aggregation or other unexpected changes of morphology. Thus, building a protective layer onto the PC₆₀BM is necessary to interrupt delamination by, in effect, stuffing the crack between the PC₆₀BM layer and the perovskite.

Concerning the power conversion efficiency (PCE) of the PSCs, it is critically related to the type of interlayer such as the electron transport layer (ETL) and hole transport layer (HTL) which have different HOMO–LUMO levels and charge carrier affinities. The effects of various interlayers are important not only for charge transport but also with regard to preventing exposure to humidity or degradation of the perovskite solar cells in atmospheric conditions.^15,16 From previous research, improved PCEs were obtained from optimized thicknesses of NiO_x, PEDOT:PSS, spiro-OMeTAD, V₂O₅ etc., as the HTL, and TiO₂, PFN, PC₆₀BM, etc., as the ETL.^17,18

In this work, we have demonstrated improved stability of PSCs by the insertion of a TiO_x ETL layer as a morphology-fixing agent for PC₆₀BM, which critically functions to block the delamination of PC₆₀BM from the perovskite layer, thereby reducing the direct penetration of water or oxygen into the perovskite layer. In the case of the MAPbI₃/PC₆₀BM bilayer, morphology changes of the PC₆₀BM during solar cell operation induce severe device degradation. By introducing a very thin morphology fixing agent, a TiO_x layer on the PC₆₀BM, the morphological change of PC₆₀BM was greatly retarded and that brought about enhanced durability of perovskite active layer. Moreover, improved electrical parameters of the PSCs with the TiO_x layer have been confirmed by the analysis of the morphology and device characteristics.

Experimental

Material preparation

CH₃NH₃I (Dyesol) and PbI₂ (Sigma Aldrich) (1 : 1 mol%) for the CH₃NH₃PbI₃ solution (35 wt%) were dissolved and stirred in a mixture of GBL (γ-butyrolactone, Sigma Aldrich) and DMSO (dimethyl sulfoxide, Junsei) (7 : 3 v/v) at room temperature for 12 h. PC₆₀BM (2.6 wt%) was also stirred in chlorobenzene (CB) at 60 °C for 12 h. Titanium(VI) isopropoxide (Sigma Aldrich) was diluted and stirred with isopropanol to spin-coating of the thicknesses from 0 to 20 nm.

Device fabrication

ITO-coated glass substrates were washed using dishwasher detergents, and then they were put into deionized water, acetone, and isopropanol during ultrasonication. PEDOT:PSS solution (AI 4083) was spin-coated on to ITO/glass at 5000 rpm for 40 s which produced a film thickness of ∼35 nm after treatment with UV ozone (20 min) and then annealing at 140 °C for 10 min. A solution of CH₃NH₃PbI₃ was coated onto ITO-glass/PEDOT:PSS substrate by spin coating to achieve a thickness of 190–290 nm. After the spin coating had been started for 15 s, the substrate was washed by dripping CB onto the substrate. After end of spin coating, the substrate was dried on a hot plate at 100 °C for 15 min. A solution of PC₆₀BM in CB was spin-coated onto ITO-glass/PEDOT:PSS/CH₃NH₃PbI₃ substrate at 1300 rpm for 60 s which produced a layer ∼100 nm thick. And then the TiO_x solution was spin-coated onto the PC₆₀BM substrate, ∼10 nm after 5000 rpm for 40 s. Finally, an Al cathode with a thickness of 100 nm was thermally deposited under 2.9 × 10⁻⁶ Torr using an evaporator. All the cells were fabricated with the average humidity being 55% (±4%).

Characterization

The light source was a 100 mW cm⁻² global AM 1.5 solar simulator that was calibrated using silicon reference cells. The J–V characteristics of the PSCs were measured using a ZIVE SP1. All the cell areas were 15 mm². The EQE was measured after power calibration (ABET 150 W xenon lamp, 13014) with a monochromator (DONGWOO OPTORN Co., Ltd., MonoRa-500i).

Other characterization methods

The FT-IR spectra were obtained using an IFS 66 V/S spectrometer and a HYPERION 3000 FTIR microscope (Bruker Optics Corporation). AFM images of several layers were taken in the noncontact mode using a Park NX10 AFM. The SEM images were acquired using a SIGMA model from Carl Zeiss, Inc. at 20 kV.

Results and discussion

Fabrication and properties of perovskite active layer with morphology fixing agent for PC₆₀BM and device performances

We fabricated perovskite solar cells (PSCs) of a planar type as shown in Fig. 1(a). The device structure was: an ITO-glass substrate, a hole transport layer (PEDOT:PSS), a perovskite layer (CH₃NH₃PbI₃), an electron transport layer (PC₆₀BM/TiO_x), and a metal electrode (Al). Fig. 1(b) shows a band energy level diagram. The CH₃NH₃PbI₃ layer was formed through a solvent-washing technology.¹⁹ As shown in Fig. 1(c), the solution-engineering process contains four steps.

Fig. 1 (a) Cell structure of the PSCs with a TiO_x layer; (b) schematic energy level diagram of each layer of ITO/PEDOT:PSS/MAPbI₃/PC₆₀BM/TiO_x/Al; (c) scheme of the fabrication method of deposition with the chlorobenzene solvent washing process during the spin coating of MAPbI₃.

First, a solution of MAPbI₃ in GBL and DMSO is dropped onto a substrate of ITO/PEDOT:PSS. Secondly, to form the layer of MAPbI₃ the spin-coater is operated and the rotation is retained so as to evaporate the solvent for 60 seconds onto the substrate, which is entirely covered by the perovskite solution. The third step is to drop chlorobenzene is on the perovskite film to wash out the GBL solvent that interrupts MAPbI₃ crystallization during spin coating. Thereby the composition of the remaining DMSO and MAPbI₃ is fixed at individual positions. The last step is an annealing procedure at 100 °C for 15 min, after which the perovskite film has become a sheet of uniform crystal structures.

In the perovskite solar cells, the grain sizes and crystallinity of the perovskite layer are crucial factors that enhance the power conversion efficiency by influencing a decrease of resistance and improved charge transfer.^20–23 In order to confirm the factors, we implemented an analysis of the morphology and crystallinity of perovskite films, and took a cross-sectional profile of the film with a scanning electron microscope (SEM).

As shown in Fig. 2(a) and (b) and S1,† the MAPbI₃ perovskite active layer is well fabricated with an average crystalline grain size of ∼150 nm. The adaptable process of the annealing condition is helpful for the fast formation of MAPbI₃ perovskite, as it is related to the uniformly crystallized grain size. We found that the films are fully-covered with a MAPbI₃ surface having a thickness of ∼190 nm and without pinholes. These results demonstrate that the MAPbI₃ film was successfully fabricated by solvent washing technology and the annealing process.

Fig. 2 SEM surface (left) and cross-sectional (right) images: (a and b) is a MAPbI₃ film, (c and d) is a coated PC₆₀BM layer on top of the MAPbI₃ film, (e and f) is a coated TiO_x layer on top of the PC₆₀BM/MAPbI₃, and (g) and (h) are proposed device schematic images which are consistent with the SEM of (c) and (e), respectively.

The deposited PC₆₀BM on top of the perovskite layer is shown in the SEM images of Fig. 2(c) and (d) and S2.† Even though the PC₆₀BM thickness of the PSCs was confirmed to be ∼100 nm, which sufficiently covered the bottom layer, there are some narrow chinks in which the material seems to be similar to individual nanoscale grain-like domains, as are observed in Fig. 2(d). We then assumed that the phenomenon of some narrow chinks occurring in the PC₆₀BM was caused by delamination between the PC₆₀BM and the perovskite. As the initial coating, the PC₆₀BM layer seems to be well fabricated on top of the perovskite surface. However, after the gradual solvent evaporation from the thin film, the adhesive force between the PC₆₀BM and perovskite is weaker than the PC₆₀BM's self-cohesiveness, which can cause aggregation and a delamination region within the layer. From this reason, we can expect that irregular PC₆₀BM region could not completely perform the function of charge transport that can enhance PCE and barrier action that can protect the MAPbI₃ layer against H₂O and O₂. Since the active layer of MAPbI₃ are easily decomposed due to the delamination of PC₆₀BM from the humid condition by previous reports,^16,24–28 we inserted a TiO_x so called morphology fixing agent between the PC₆₀BM and metal electrode to prevent the delamination and to induce cascaded charge transport. Thereby the TiO_x will be able to contribute to prohibit unexpected morphological changes and degradation of the exposed MAPbI₃/PC₆₀BM layer as shown in proposed schematic of Fig. 2(g) and (h).

To confirm the chemical components of MAPbI₃/PC₆₀BM with a thin layer of the TiO_x (approximately ∼10 nm),^29–31 we performed FT-IR analysis as shown in Fig. 3, and identified the main four vibrational lines in the spectra of the layers. Two peaks were observed at 3182 cm⁻¹, and 1465 cm⁻¹ which were respectively matched to the NH₃⁺ stretch, and CH₃ bend.³² Additional meaningful peak at 1736 cm⁻¹ was assigned to an ester such as –RCOOCH₃ at end of an alkyl chain of PC₆₀BM. And the other peak at 1182 cm⁻¹ was assigned to the Ti–O–C stretching mode in the TiO_x layer.³³ According to these observations, we found that the TiO_x were well covered without interfacial decomposition which preserve the chemical component of MAPbI₃/PC₆₀BM layer during the sequential deposition.

Fig. 3 The results of the FT-IR analysis: (a) MAPbI₃, (b) MAPbI₃/PC₆₀BM and (c) MAPbI₃/PC₆₀BM/TiO_x.

In PSCs, to obtain enhanced device performance, the thicknesses of the MAPbI₃ and PC₆₀BM layers should be optimized by controlling the revolutions per minute (rpm) during the spin coating process. A relatively thin layer of PC₆₀BM (below ∼100 nm)^34,35 was initially coated on top of an MAPbI₃ layer of optimized thickness, as thick film brings about a decreased FF as shown in Fig. S3 and Table S1.† After that, we conducted a much more detailed experiment of device optimization by changing the thicknesses of TiO_x layer as shown in Fig. S4 and Table S2.† Here, the best device performance exhibits observed at thickness ∼10 nm of TiO_x due to the increased FF and J_SC. Because the thicker interlayer tends to increase series resistance at the device and thinner device have little effect due to the low coverage, therefore thickness of morphology fixing agent is very important for optimized PCE.

Fig. 4(a) and Table 1 shows the J–V curves of the devices depending on the various thicknesses of PC₆₀BM and without or with a TiO_x layer added onto the PC₆₀BM. The device fabricated with only PC₆₀BM has a PCE of 8.01% with a V_OC of 0.875 V, a J_SC of 12.58 mA cm⁻², and a FF of 72.8%. The device with the optimized interlayer PC₆₀BM/TiO_x exhibited an improved PCE of 10.41% with a V_OC of 0.928 V, a J_SC of 14.78 mA cm⁻², and a FF of 75.9%.

Fig. 4 (a) J–V curves of the devices fabricated as MAPbI₃/PC₆₀BM depending on the different thicknesses and the presence or absence of the fixing agent of the TiO_x layer, under the AM 1.5G; (b) dark J–V curves of the PSCs without and with the TiO_x layer.

Table 1 Electrical parameters of the PSCs depending on the controlled thickness of PC₆₀BM and PSCs with a TiO_x layer. The thickness of the MAPbI₃ layer is optimized at 190 nm

Thickness of PC60BM	VOC (V)	JSC (mA cm^−2)	FF (%)	Eff. (%)
130 nm	0.859	11.44	76.3	7.49
100 nm	0.875	12.58	72.8	8.01
70 nm	0.844	13.18	70.2	7.81
100 nm with TiOx	0.928	14.78	75.9	10.41

---

The increased electrical parameters of the PSCs are related to the role of the TiO_x layer as a morphology fixing agent, which might be the origin of a reduced device series resistance R_s and significant improvement of charge carrier transport. We also checked the J–V curves of the device without and with TiO_x, by using a reverse scan for hysteresis as shown in Fig. S5 and Table S3.† Moreover, we measured J–V curve in the dark and then, the shunt resistance R_sh and also R_s were obtained from Fig. 4(b). As shown in Table 1, the values of R_sh and R_s were respectively 2.03 kΩ cm² and 7.66 Ω cm² in the device without TiO_x but were 76.45 kΩ cm² and 3.18 Ω cm² in the cell with TiO_x. Therefore, the FF is favorably enhanced from 72.8% to 75.9% in the device with TiO_x due to the lower series resistance and increased shunt resistances.

Analysis of charge transport and stability in perovskite solar cells with morphology fixing agent for PC₆₀BM

Moreover, to characterize the effect of TiO_x layer for the enhancement of the J_SC, we measured and analyzed the EQE, the space-charge-limited current (SCLC), and the photocurrent density J_ph as a function of the internal voltage V_int, as shown in Fig. 5. As shown in Fig. 5(a), the overall range of the improved external quantum efficiency (EQE) is observed to occur at 300–780 nm by comparing the devices without and with TiO_x. The measured J_SC and calculated J_SC of the PSCs without (12.58 mA cm⁻² vs. 12.06 mA cm⁻²) and with TiO_x layer (14.78 mA cm⁻² vs. 14.19 mA cm⁻²) revealed considerably similar with the error range of less than 4%. So we surmise that the TiO_x layer favors light absorption at whole range of 300–780 nm. To confirm the difference of hole mobility without and with TiO_x layer, we observed the influence of the SCLC by the measuring J–V curves.^36,37 In high voltage range, the J–V characteristics of the devices can be coherently represented by the Mott–Gurney This equation can be expressed as where ε_r is the relative permittivity of the medium, ε₀ is the permittivity of the vacuum and L is the thickness of the perovskite active layer. By substitution in the formula, the hole mobility without the TiO_x layer was computed at 9.31 × 10⁻⁵ cm² V⁻¹ s⁻¹ but with the TiO_x layer it was 3.51 × 10⁻⁴ cm² V⁻¹ s⁻¹, as in Fig. 5(b). With the insertion of the TiO_x layer in PSCs, we demonstrated that the hole mobility was significantly improved over the PSCs without the TiO_x.

Fig. 5 (a) External quantum efficiency; (b) hole mobility; and (c), the photocurrent density J_ph versus internal voltage V_int characterization of the PSCs without and with a TiO_x layer.

To characterize and compare the properties of charge generation and extraction in the PSCs without and with TiO_x, we analyzed the photocurrent density J_ph with the internal voltage of V_int. The J_ph was defined to be J_ph = J_L − J_D, where J_L and J_D are the current densities under light and in the dark respectively. So, the V_int was determined by V_int = V₀ − V_apl, where V₀ is the voltage at which J_ph is zero and V_apl is the applied voltage.^38,39 Fig. 5(c) is the result of J_ph versus V_int. J_ph saturated at high V_int. Thus, we assume that most of the excitons that were generated by photons are separated into free charge carriers at high V_int over 2 V. We found that the PSCs with TiO_x showed a higher saturation J_ph (13.02 mA cm⁻²) than that of those without (11.73 mA cm⁻²) at V_int ∼ 2 V. So charge generation in this device was more efficient than in the pristine device. Also, we observed an obvious improvement of J_ph at low V_int in the PSC with TiO_x and the higher J_ph shows that charge extraction efficiency was generally enhanced. A higher charge extraction ability plays a role in improving the efficiency of the device. Overall Fig. 5 shows that the improved EQE, hole mobility, and charge generation and extraction contributed to the higher J_SC with the TiO_x layer than without it (see the summary of values in Table 2).

Table 2 Hole mobility and J_ph of the PSCs without TiO_x (pristine) and with the TiO_x

PSCs	Hole mobility (cm^2 V^−1 s^−1)	Jph (mA cm^−2) (at Vint = 2 V)
Without TiOx	9.31 × 10^−5	11.73
With TiOx	3.51 × 10^−4	13.02

---

From the above information, we surmise that PC₆₀BM does not function perfectly as a charge transport and barrier layer. The phenomenon was revealed by observations made during a stability test of the PCE as a function of time, as shown in Fig. 6.

Fig. 6 Device durability test: cell efficiency dependence on then number of days from fabrication, without and with the morphological fixing agent TiO_x.

The pristine PSCs of CH₃NH₃PbI₃ are sharply decomposed at a relatively normal humidity owing to the hygroscopic amine salts.¹⁶ The PSCs without TiO_x showed as much as a 50% reduction of the PCE after only one day. But the PSCs that incorporated TiO_x took 13 days to exhibit a 50% decrease. Some regions of the perovskite layer were directly exposed to the ambient atmosphere because of incompletely covering by the PC₆₀BM that was formed with some island formations. However the TiO_x layer works as not only morphology-fixing component, but also water-proof effect that blocks delamination and combines atomic with oxygen in H₂O and O₂, bringing about longer device operation.^33,36 When the cells were fabricated, immediately each of the PSCs seemed to have a very clear surface. But a severe degradation region was formed around the aluminum electrode after 3 days, especially on the PSCs without the TiO_x layer on which there was much faster metal electrode oxidation, as shown in Fig. 6 and S6.†

To verify the degradation morphology depending on time, we checked the SEM images of the perovskite, perovskite/PC₆₀BM, and, perovskite/PC₆₀BM/TiO_x. In Fig. 7, the layer of perovskite exhibited unstable morphology after 7 days due to the existence of pinhole and defect sites, which bring about sharply increased rms roughness from 7.7 nm to 13.1 nm in the AFM analysis and the SEM results of Fig. S7.† PC₆₀BM film revealed unstable surface morphology owing to sharply increased rms from 2.8 nm to 32.3 nm by PC₆₀BM delamination as shown in Fig. 7(c) and (d). In the other hands, there is shown a relatively stable and retained surface morphology of the device with TiO_x layer due to the remaining surface of rms from 1.4 nm to 2.9 nm with the effect of morphology fixing agent (see Fig. 7(e) and (f)). Finally, Fig. 8 shows the histograms of the average PCEs for all of independently fabricated cells that contributed to our study. The PCE tendency shows that the PC₆₀BM/TiO_x cascaded interlayer approaches over average 10% compared to the pristine device.

Fig. 7 AFM height images of the (a) MAPbI₃, (c) MAPbI₃/PC₆₀BM, (e) MAPbI₃/PC₆₀BM/TiO_x; (b), (d) and (f) are the same active layer after 7 days with increased surface roughness rms values compared to those of the pristine samples. The overall roughness trend shows MAPbI₃/PC₆₀BM > MAPbI₃ > MAPbI₃/PC₆₀BM/TiO_x.

Fig. 8 The tendency of the cell efficiency: pristine PSCs versus PSCs with a TiO_x layer.

Conclusions

In conclusion, we have demonstrated that the stability of PC₆₀BM morphology on a perovskite layer is crucial for not only initial device efficiency, but also for long-term performance and reliability. Unexpectedly, it was found that a thin TiO_x layer which was originally inserted to improve the efficiencies of perovskite solar cells (PSCs) could act as a morphology-fixing agent and water-proof effect during long-term device operation. We can summarize two main effects of the TiO_x in PSCs. First, the TiO_x layer plays the important role of increasing device electrical parameters such as the J_SC, FF, and the V_OC due to the improvements in EQE, resistance, and charge extraction/generation that were shown via J_ph versus V_int characterization and SCLC measurements of mobility. Secondly, the device operation exhibited relatively long-term durability due to stabilizing the morphology and covering the cracks in the PC₆₀BM layer, and also hindering unwanted surface changes at the interface between the perovskite and PC₆₀BM layers. Due to the combination of those factors, the morphology fixing agent is a remarkable strategy for enhancing the efficiency, reliability and stable operation of PSCs.

Acknowledgements

This research was supported by the Basic Science Research Program, through the National Research Foundation of Korea (NRF), funded by the Ministry of Science, ICT & Future Planning (2014R1A1A1002419, NRF-2015M2A2A6A01045277, 2014M3A7B4052200).

Notes and references

1. H. J. Snaith, J. Phys. Chem. Lett., 2013, 4, 3623–3630.
2. G. E. Eperon, S. D. Stranks, C. Menelaou, M. B. Johnston, L. M. Herz and H. J. Snaith, Energy Environ. Sci., 2014, 7, 982–988.
3. T. C. Sum and N. Mathews, Energy Environ. Sci., 2014, 7, 2518–2534.
4. S. D. Stranks, G. E. Eperon, G. Grancini, C. Menelaou, M. J. P. Alcocer, T. Leijtens, L. M. Herz, A. Petrozza and H. J. Snaith, Science, 2013, 342, 341–344.
5. 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.
6. Y. Zhao and K. Zhu, J. Phys. Chem. Lett., 2013, 4, 2880–2884.
7. C. Wehrenfennig, G. E. Eperon, M. B. Jonston, H. J. Snaith and L. M. Herz, Adv. Mater., 2014, 26, 1584–1589.
8. J. H. Heo, D. H. Song and S. H. Im, Adv. Mater., 2014, 26, 8179.
9. N. G. Park, J. Phys. Chem. Lett., 2013, 4, 2423–2429.
10. H. Zhou, Q. Chen, G. Li, S. Luo, T. B. Song, H. S. Duan, Z. Hong, J. You, Y. S. Liu and Y. Yang, Science, 2014, 345, 542–546.
11. S. Das, B. Yang, G. Gu, P. C. Joshi, I. N. Ivanov, C. M. Rouleau, T. Aytug, D. B. Geohegan and K. Xiao, ACS Photonics, 2015, 2, 680–686.
12. Y. H. Chen, T. Chen and L. Dai, Adv. Mater., 2015, 27, 1053–1059.
13. M. Z. Liu, M. B. Johston and H. J. Snaith, Nature, 2013, 501, 395–398.
14. D. Liu, M. K. Gangishetty and T. L. Kelly, J. Mater. Chem. A, 2014, 2, 19873–19881.
15. Y. J. Jeon, S. H. Lee, R. R. Kang, J. E. Kim, J. S. Yeo, S. H. Lee, S. S. Kim, J. M. Yun and D. Y. Kim, Sci. Rep., 2014, 4, 6953.
16. J. H. Noh, S. H. Im, J. H. Heo, T. N. Mandal and S. I. Seok, Nano Lett., 2013, 13, 1764–1769.
17. P. Docampo, J. M. Ball, M. Darwich, G. E. Eperon and H. J. Snaith, Nat. Commun., 2013, 4, 2761.
18. N. J. Jeon, J. Lee, J. H. Noh, M. K. Nazeeruddin, M. Gratzel and S. I. Seok, J. Am. Chem. Soc., 2013, 135, 19087–19090.
19. N. J. Jeon, J. H. Noh, Y. C. Kim, W. S. Yang, S. C. Ryu and S. I. Seok, Nat. Mater., 2014, 13, 897–903.
20. J. Y. Jeng, Y. F. Chiang, M. H. Lee, S. R. Peng, T. F. Guo, P. Chen and T. C. Wen, Adv. Mater., 2013, 25, 3727–3732.
21. G. E. Eperon, V. M. Burlakov, P. Docampo, A. Goriely and H. J. Snaith, Adv. Funct. Mater., 2014, 24, 151–157.
22. 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.
23. Y. Huang, C. Tsao, Y. Cho, K. Chen, K. Chiang, S. Hsiao, C. Chen, C. Su, U. Jeng and H. Lin, Sci. Rep., 2015, 5, 13657.
24. P. M. Buschbaum, Adv. Mater., 2014, 26, 7692–7709.
25. X. Dong, X. Fang, M. Lv, B. Lin, S. Zhang, J. Ding and N. Yuan, J. Mater. Chem. A, 2015, 3, 5360–5367.
26. S. N. Habissreutinger, T. Leijtens, G. E. Eperon, S. D. Stranks, R. J. icholas and H. J. snaith, Nano Lett., 2014, 14, 5561–5568.
27. J. Xu, A. Buin, A. H. Ip, W. Li, O. Voznyy, R. Comin, M. Yuan, S. Jeon, Z. Ning, J. J. McDowell, P. Kanjanaboos, J. P. Sun, X. Lan, L. N. Quan, D. H. Kim, I. G. Hill, P. Maksymovych and E. H. Sargent, Nat. Commun., 2015, 6, 7081.
28. A. H. Ip, L. N. Quan, M. M. Adachi, J. J. McDowell, J. Xu, D. H. Kim and E. H. Sargent, Appl. Phys. Lett., 2015, 106, 143902.
29. D. H. Wang, S. H. Im, H. K. Lee, O. O. Park and J. H. Park, J. Phys. Chem. C, 2009, 113, 17268–17273.
30. K. Lee, J. Y. Kim, S. H. Park, S. H. Kim, S. Cho and A. J. Heeger, Adv. Mater., 2007, 19, 2445–2449.
31. J. K. Kim, W. Kim, D. H. Wang, H. Lee, S. M. Cho, D. G. Choi and J. H. Park, Langmuir, 2013, 29, 5377–5382.
32. T. Glaser, C. Muller, M. Sendner, C. Krekerler, O. E. Semonin, T. D. Hull, O. Yaffe, J. S. Owen, W. kowalsky, A. Pucci and R. Lovrincic, J. Phys. Chem. Lett., 2015, 6, 2913–2918.
33. R. Urlaub, U. Posset and R. Thull, J. Non-Cryst. Solids, 2000, 265, 276–284.
34. C. R. Carmona, O. Malinkiewicz, A. Soriano, G. M. Espallargas, A. Garcia, P. Reinecke, T. Kroyer, M. I. Dar, M. K. Nazeeruddine and H. J. Bolink, Energy Environ. Sci., 2014, 7, 994–997.
35. S. Sun, T. Salim, N. Mathews, M. Duchamp, C. Boothroyd, G. Xing, T. C. Sum and Y. M. Lam, Energy Environ. Sci., 2014, 7, 399–407.
36. Z. He, C. Zhong, X. Huang, W. Y. Wong, H. Wu, L. Chen, S. Su and Y. Cao, Adv. Mater., 2011, 23, 4636–4643.
37. C. Melzer, E. J. Koop, V. D. Mihailetchi and P. W. Blom, Adv. Funct. Mater., 2004, 14, 865–870.
38. A. K. K. Kyaw, D. H. Wang, D. Wynands, J. Zhang, T. Q. Nguyen, G. C. Bazan and A. J. Heeger, Nano Lett., 2013, 13, 3796–3801.
39. S. Cho, K. Lee and A. J. Heeger, Adv. Mater., 2009, 21, 1941–1944.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra08584a

‡ S. Ahn and W. Jang contribute equally in this work.

---