Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

# Strategic advantages of reactive polyiodide melts for scalable perovskite photovoltaics

## Abstract

Despite tremendous progress in efficiency and stability, perovskite solar cells are still facing the challenge of upscaling. Here we present unique advantages of reactive polyiodide melts for solvent- and adduct-free reactionary fabrication of perovskite films exhibiting excellent quality over large areas. Our method employs a nanoscale layer of metallic Pb coated with stoichiometric amounts of CH3NH3I (MAI) or mixed CsI/MAI/NH2CHNH2I (FAI), subsequently exposed to iodine vapour. The instantly formed MAI3(L) or Cs(MA,FA)I3(L) polyiodide liquid converts the Pb layer into a pure perovskite film without byproducts or unreacted components at nearly room temperature. We demonstrate highly uniform and relatively large area MAPbI3 perovskite films, such as 100 cm2 on glass/fluorine-doped tin oxide (FTO) and 600 cm2 on flexible polyethylene terephthalate (PET)/indium tin oxide (ITO) substrates. As a proof-of-concept, we demonstrate solar cells with reverse scan power conversion efficiencies of 16.12% (planar MAPbI3), 17.18% (mesoscopic MAPbI3) and 16.89% (planar Cs0.05MA0.2FA0.75PbI3) in the standard FTO/c(m)-TiO2/perovskite/spiro-OMeTAD/Au architecture.

## Main

The meteoric rise of the power conversion efficiencies (PCE) of perovskite solar cells (PSCs) from 3.8% to over 23% in less than a decade has made them superstars and stunned the photovoltaic community1,2,3,4. Techno-econometric models predict a significant cost reduction potential for perovskite photovoltaic (PV) modules5. Strong expectations are also put on perovskite tandems6,7,8 that could overcome the Shockley–Queisser PCE limit of 33.5%9. Although the issues of ecotoxicity10, hysteresis11 and stability12,13 are progressively being addressed, the remaining most important challenge is the development of truly scalable fabrication methods of perovskite absorber layers. PSCs undergo a noticeably larger decrease of PCE on scaling than other PV technologies14. A comparison of potentially scalable solvent-15,16,17,18,19,20,21,22,23,24,25, amine-26,27,28,29,30,31 and vacuum-assisted32,33 methods is beyond the scope of this paper, but we recommend the reviews14,34 on scalability issues of perovskite photovoltaics in addition to our Supplementary Discussion 1. However, we do want to stress here that, despite occasional demonstration of modules with promising efficiencies15,16,17,35, no deposition method so far has demonstrated evidence of robust scalability with a potential to trigger commercialization of PSCs. For that reason, the development of novel fabrication methods that can open up new directions towards readiness of perovskite PV technology for mass-scale manufacturing is of utmost importance.

From the fundamental point of view, the great variety of the reported fabrication methods can be grouped into two categories:

1. 1.

Crystallization from extrinsic media: MAPbI3(Sol) → MAPbI3(S) + Sol (that is, from solutions in polar aprotic solvents2,15,16,17 or liquid amine solvates26,27,28,29,30,31)

2. 2.

Combination reaction of two halides: PbI2(S/V) + MAI(S/Solv/V) → MAPbI3(S) (that is, dipping of PbI2 films into MAI solution in isopropanol (IPA)36, annealing of PbI2 in MAI vapour by hybrid CVD33, co-evaporation of PbI2 and MAI32, and so on)

These methods are implemented in many different flavours; however, in none of them are hybrid perovskite films obtained through a direct redox reaction. In this work, we present a fundamentally different method that is based on a direct redox reaction between nanoscale layers of metallic Pb and liquid polyiodides: MAI3(L) or mixed MA(Cs,FA)I3(L). Here, the change of the oxidation states of Pb0 → Pb2+ and I3 → 3I creates a strong driving force for the direct conversion of metallic Pb into highly crystalline and uniform perovskite layers. Being fundamentally different, our polyiodide-assisted method is not affected by the limitations of solvent- and vacuum-assisted methods and thus opens up a distinct branch of technologies with their own strategic advantages in the field of perovskite PV.

Liquid polyiodides with the general formula AIx (A: organic/inorganic cation) were known for many AI–I2 systems, where AI is organic R3SI, R4NI (R: methyl, ethyl, n-propyl, n-butyl and so on) and inorganic CsI, NH4I iodides37. However, liquid MAI3 and FAI3, and their high reactivity towards Pb, leading to the formation of perovskite films, have recently been discovered by Tarasov et al.38, who coined their name as reactive polyiodide melts (RPM). Although the formation of highly crystalline MAPbI3 films has been demonstrated by spin coating of liquid MAI3 over Pb layers38, an excess of highly reactive MAI3 frequently caused dissolution and recrystallization of the MAPbI3 films. To obtain a 500-nm-thick film of MAPbI3, a 62 nm layer of Pb should react stoichiometrically with a 230 nm layer of liquid MAI3, which is beyond the range of reproducible control for spin or slot-die coating.

In a striking contrast to the previous work focused on the reaction of Pb with MAI3(L) obtained by mixing of two solid precursors (that is, MAI(S) + I2(S) → MAI3(L))38, we discovered that the MAI3(L) can be obtained through a simple interaction of solid MAI with iodine vapour (that is, MAI(S) + I2(V) → MAI3(L)). This enabled an entirely novel synthetic approach to the fabrication of perovskite thin films via controllable, vapour-phase-triggered, formation of nanoscale MAI3(L) layer in stoichiometric quantities to the underlying Pb layer starting with Pb/MAI bilayer (that is, Pb(S)/MAI(S) + I2(V) → Pb(S)/MAI3(L) → MAPbI3(S)).

## Nanoscale reactive polyiodide layers and perovskite films

Figure 1a illustrates the proposed method of reactive polyiodide melt-assisted growth through in-situ conversion (RP-MAGIC) of Pb into hybrid perovskite. The stoichiometric amounts of Pb and MAI in the initial bilayer ensure that there are neither unreacted components nor byproducts remaining after the reaction, since all species involved in the process are consumed entirely, thus constituting the final perovskite film. To demonstrate scalability of our method, we fabricated MAPbI3 films with a large grain morphology and outstanding uniformity on 10 × 10 cm2 glass/FTO (Fig. 1b and Supplementary Fig. 1) and 20 × 30 cm2 flexible PET/ITO (Fig. 1c, Supplementary Fig. 2 and Supplementary Video 1) substrates.

We have studied the conversion process in situ by measuring X-ray diffraction (XRD) from a glass/Pb/MAI sample located near a small I2 crystal in an Ar-filled capsule (Supplementary Fig. 3). Iodine is moderately volatile, with vapour pressures of 37–82 Pa at 25–35 °C. A small crystal of iodine placed near a sample becomes a sufficient source of iodine vapour to initiate conversion of the Pb/MAI bilayer into MAPbI3 (Supplementary Fig. 4 and Supplementary Video 2). Figure 2a,b proves the gradual disappearance of MAI (021) and Pb (111) peaks and the rise of MAPbI3 (110) peaks, revealing the direct conversion of the Pb/MAI bilayer into MAPbI3 without the formation of intermediate phases. We attribute a rather weak PbI2 (001) peak emerging at the final stages of the process to air-induced degradation of the MAPbI3 in the slowly leaking capsule, as it does not appear for the samples converted in an inert glove-box. Although exposure to I2 vapour has been reported to cause MAPbI3 degradation39, we do not observe such phenomena under our experimental conditions. In fact, Wang et al.39 conducted their studies with high iodine source temperatures of 90 °C that created an extremely high partial vapour pressure of iodine above 3,300 Pa, which is almost 100 times higher than in our case and probably beyond the stability region of MAPbI3.

We have used the Raman spectroscopy (Fig. 2c) to confirm the existence of the liquid MAI3 phase during the conversion process. Initially, no reaction occurs between Pb and MAI before iodination and only a known MAI vibration mode at 115 cm−1 is present. The typical spectrum of MAI338 appears only after exposure to I2 vapour, with characteristic vibrations near 110 cm−1 and 160 cm−1 that correspond to an I3 symmetrical stretch vibration and to I2 molecular units solvated with I3, respectively37.

Whereas a good uniformity of nanoscale Pb layers is essential for the RP-MAGIC method, it is less important for MAI, which liquefies into MAI3 (Supplementary Fig. 5) followed by uniform wetting and spreading over the Pb layer. Indeed, we obtained MAPbI3 films with an excellent morphology from the Pb/MAI bilayer with a rather rough MAI capping fabricated by spray deposition (Supplementary Fig. 6).

The growth of perovskite films from solutions is accompanied by volume shrinking of coated liquid layers upon solvent evaporation and intermediate adduct decomposition, which becomes the primary source of pinholes in the solution-processed films. In contrast, the combination reactions: PbI2(S) + MAI(V/L/S) → MAPbI3(S) and Pb(S) + MAI3(L) → MAPbI3(S) lead to the 2 and 8 times volume expansion of the Pb precursor layers, respectively, on their conversion into MAPbI3 that promotes healing of pinholes. However, the weak driving force for the combination reaction of PbI2 with MAI (refs 33,36) frequently leads to diffusion-limited incomplete conversion of PbI2, which is also true for the MAI reaction with Pb (ref. 40), PbCl2, PbO, Pb(CH3COO)2(ref. 41), that results in formation of MAPbI3 via intermediate PbI2. In contrast, the redox reaction of metallic Pb with liquid MAI3 creates a strong driving force for the conversion process. In combination with the high volume expansion (Fig. 3a,b), it promotes the formation of large perovskite grains already at room temperature. Also, there is no compromise on films purity, because extrinsic substances are not involved in the process. Considering how easily Pb layers can be fabricated in a highly uniform manner by a vacuum deposition, the RP-MAGIC method shows great promise to become an ultimate solution for manufacturing of pinhole-free perovskite layers with exceptional crystalline and optoelectronic quality over large areas. Indeed, the optical absorption, steady-state and time-resolved photoluminescence show negligible variation across MAPbI3 films fabricated on large 10 × 10 cm2 glass substrates (Fig. 3e–i). The map of optical absorption measured at λ = 740 nm varies in the range of only ± 5% (Supplementary Fig. 1d). Time-resolved photoluminescence (TRPL) reveals long charge-carrier lifetimes of τ1 = 2229 ns and τ2 = 188197 ns within the double-exponential decay model. Considering electron and hole mobilities for MAPbI3 in the range of 2.525 cm2 V−1 s−1(refs 42,43), the carrier diffusion length $$L_{\mathrm{D}} = \sqrt {\tau \mu kT/q}$$ (where μ is mobility, τ is charge-carrier lifetime, k is Boltzmann constant, q is elementary charge and T is temperature) exceeds the film thickness of 500 nm already for the short decay component. Thus there are no fundamental obstacles for obtaining solar cells with remarkable PCE by our method.

## Conversion mechanisms and multi-component phases

According to the working phase diagram (Fig. 4a), there are several possible outcomes of the conversion process, depending on relative amounts of Pb and MAI in the bilayer and iodine intake (see Supplementary Fig. 7 and Supplementary Discussion 2). In the case of a low partial vapour pressure of iodine, the slow iodine intake MAI + I2 → MAIz becomes a rate limiting step and ensures successful transformation of the stoichiometric Pb/MAI bilayer into the pure MAPbI3 film upon the reaction of MAIz=3 with Pb. In contrast, the rapid intake of iodine at high iodine vapour pressures can lead to quick enrichment of the polyiodide phase to MAIz>6 compositions, resulting in the formation of PbI2. Although an excess of Pb in the Pb/MAI bilayer could lead to Pb/MAPbI3 or PbI2/MAPbI3, depending on the iodination time, a reasonable excess of MAI could be easily tolerated until the enrichment of the residual MAI3(L) with iodine is prevented. For instance, such residual MAI3(L) can be gently transformed back to MAI at 50–60 °C and then washed away in isopropyl alcohol (IPA). Thus, a rather robust process window for the RP-MAGIC method is ensured by the tolerance to MAI excess and different possibilities for the control of the iodine vapour treatment.

Furthermore, we determined that all polyiodide phases related to the technologically attractive mixed-cation and mixed-halide perovskites44,45 have melting points below 65 °C (Fig. 4b), thus proving the general applicability of the RP-MAGIC method within the relevant compositional space. Consequently, we have fabricated technologically promising triple-cation Cs0.05MA0.2FA0.75PbI3(ref. 45) films from multi-layer Pb/(CsI)0.05(FAI)0.75(MAI)0.2 stacks (Supplementary Fig. 8). The XRD pattern (Fig. 5b) confirms the cubic lattice for the triple-cation perovskite film, while its optical bandgap (Fig. 5d) and the photoluminescence peak position (Fig. 5c) are red-shifted in comparison to the MAPbI3. The XPS studies (Supplementary Figs. 9 and 10) confirm incorporation of Cs, FA and MA cations into the film, in satisfactory agreement with the target composition.

## Performance of photovoltaic devices

To demonstrate that our method can produce efficient absorber layers, we have fabricated and characterized standard FTO/c(m)-TiO2/perovskite/spiro-OMeTAD/Au solar cells (see Supplementary Discussion 3). Our first generation of mesoscopic (M1) and planar (P1) devices was prepared by using a basic recipe without doping of TiO2 and MAPbI3 layers. Supplementary Fig. 11 demonstrates the best-performing devices from the M1 and P1 batches with PCE of 12.24% and 13.15% in reverse current–voltage (JV) scans and pronounced hysteresis of 8% and 40%, respectively. Without undertaking full optimization attempts, we improved the reverse scan PCE to 16.12% and 17.18% and reduced hysteresis to ~3% and ~1% in the second-generation planar (P2) and mesoscopic (M2) devices (Fig. 6) by using Li, Mg co-doping of c-TiO2 (ref. 46), Li doping of m-TiO2(refs 46,47) and LiTFSI, FK-209 Co-complex, TBP co-doping of the spiro-OMeTAD (ref. 44) in addition to thickness optimization and a KI post-treatment of the perovskite layers inspired by Tang et al46. In addition, the distributions of performance parameters were noticeably narrowed for both P2 and M2 batches (Supplementary Figs. 12 and 13).

Furthermore, we have fabricated planar (P3) Cs0.05MA0.2FA0.75PbI3 solar cells (Supplementary Fig. 14). The reverse scan PCE of the best P3 device (16.89%) exceeded that of the best P2 device due to improved JSC (22.88 versus 21.63 mA cm−2), but remained slightly below the PCE of the best M2 device due to the smaller open circuit voltage VOC (1.07 versus 1.10 V) and fill factor FF (0.69 versus 0.74). This actually shows a potential for further PCE improvement and we anticipate that a PCE of over 20% can be achieved after optimization of the Cs0.05MA0.2FA0.75PbI3 absorber in the mesoscopic architecture. To prove the excellent uniformity of the perovskite layers upon scaling, we prepared three planar (P4) MAPbI3 solar cells with active areas of 2.45 cm2 each on the same 5 x 5 cm2 substrate that demonstrated negligible variation of the reverse scan PCE of 14.27 ± 0.06% (Supplementary Fig. 15).

The PCE of the (P3) Cs0.05MA0.2FA0.75PbI3 and (M2) MAPbI3 devices measured under continuous illumination at the maximum power point bias voltage between JV scans during 100 hours (Supplementary Fig. 16d) retained 80% and 50% of the initial values, respectively. This decrease is typical for devices containing heavily doped spiro-OMeTAD, where Li-TFSI dopant triggers degradation of the perovskite layers12. In contrast, the Cs0.05MA0.2FA0.75PbI3 and MAPbI3 films on glass substrates show no sizeable degradation under the same illumination conditions (Supplementary Fig. 16e,f).

Here it is important to stress that our experimental results irrefutably demonstrate that our method can be used to fabricate hybrid perovskite solar cells with high efficiencies, albeit the fact that our devices are not yet fully optimized. Since our studies (SEM, XRD, PL and TRPL) confirm the high structural and optoelectronic quality of the perovskite films prepared by our method, the factors that limit the efficiency of our devices are highly likely related to the insufficient performance of the electron- and hole-transport layers. For that reason, we believe that implementation of superior electron-transport layers, such as SnO2 (ref. 48), Nb:SnO2 (ref. 49) and La:BaSnO3 (ref. 50), together with more stable hole-transport layers, such as PTAA (ref. 4) and CuSCN (ref. 13) are reliable approaches to achieve state-of-the-art performance in our solar cells.

## Conclusions

We have developed an innovative method for the fabrication of highly crystalline and uniform perovskite films over large areas that meets all the essential criteria for becoming a scalable mass-production technology for perovskite photovoltaics. Our method relies on polyiodide-assisted conversion of nanoscale layers of metallic Pb into hybrid perovskite films with exceptionally high crystalline quality already at room temperatures. Highly reactive polyiodide melts that are instantly formed upon exposure of stoichiometric Pb/MAI or Pb/CsI/FAI/MAI stack films to iodine vapour react with the underlying Pb layers and convert them into pure hybrid perovskite films without byproducts or unreacted components. The versatility, scalability and robustness of our innovative method has been demonstrated through the fabrication of MAPbI3 films on rigid 100 cm2 glass/FTO and 600 cm2 flexible PET/ITO substrates with perfect morphology and coverage. Solar cell devices with proof-of-concept power conversion efficiencies of 16.12% and 17.18% (reverse scan) have been demonstrated for the planar and mesoscopic FTO/c(m)-TiO2/MAPbI3/spiro-OMeTAD/Au device architectures, respectively, while the planar device based on a Cs0.05MA0.2FA0.75PbI3 absorber layer achieved an efficiency of 16.89% (reverse scan). These results demonstrate several strategic advantages of the RP-MAGIC approach as a potent new low-temperature melt-based perovskite solar cell production technology. First of all, a new, condensed, low-temperature melt phase serves as a self-sufficient precursor including only the natural components of the hybrid perovskite, thus successfully replacing diluted gaseous or solution reagents. Then, a fast solid–melt chemical reaction of Pb and RPM manifests the only essential step for direct and instant formation of pure perovskite without byproducts and additional microstructure-worsening stages caused by adduct formation, decomposition and the evolution of all residual extrinsic molecules that are unavoidable in conventional solvent-assisted methods. High diffusivity of components in polyiodide molten phases provides the fast supply of reagents and finally results in better crystallinity of the perovskite layers. The redox nature of the conversion reaction provides quick kinetics, easy stoichiometry control and demands no additional heating. Moreover, the reaction proceeds instantly and under isothermal conditions in a self-propagating and self-supporting manner. Furthermore, a large overall volume increase during Pb conversion into perovskite guarantees the formation of a pinhole-free layer of the perovskite. Also, an easy control over perovskite composition and thickness is achieved by depositing a sandwiched structure of Pb and single- or mixed-halide components, predetermining the final composition of perovskite layers. Importantly, the sandwiched film is an all-solid-state layered matter with kinetically and redox hindered interactions, which can be easily and simultaneously triggered at each point by gradual absorption of a needed amount of iodine from a vapour supply. This, in turn, results in excellent scalability and microstructural uniformity of the obtained perovskite films, because the forming low-temperature melt wets the underlying Pb layer well, creeps over it and forms a conformal and fluent liquid layer that redistributes further upon the reaction to grow one and the same perovskite layer at each point with a high reproducibility. The RP-MAGIC method has a wide range of opportunities for further improvements and progress, and we believe that a better, easily possible, control over the sandwiched structure thickness and composition, more precise control of the iodine supply and performance optimization of electron- and hole-transport layers would allow record efficiencies to be achieved in the near future, substantiating the present proof-of-concept.

## Methods

Methods are provided in the Supplementary Information.

### Reporting Summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.

## Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

## References

1. 1.

Kojima, A., Teshima, K., Shirai, Y. & Miyasaka, T. Organometal halide perovskites as visible-light sensitizers for photovoltaic cells. J. Am. Chem. Soc. 131, 6050–6051 (2009).

2. 2.

Yang, W. S. et al. High-performance photovoltaic perovskite layers fabricated through intramolecular exchange. Science 348, 1234–1237 (2015).

3. 3.

NREL Best Research-Cell Efficiencies; https://www.nrel.gov/pv/assets/images/efficiency-chart.png

4. 4.

Saliba, M. et al. Incorporation of rubidium cations into perovskite solar cells improves photovoltaic performance. Science 354, 203–206 (2016).

5. 5.

Chang, N. L. et al. Manufacturing cost and market potential analysis of demonstrated roll-to-roll perovskite photovoltaic cell processes. Sol. Energy Mater. Sol. Cells 174, 314–324 (2018).

6. 6.

Forgács, D. et al. Efficient monolithic perovskite/perovskite tandem solar cells. Adv. Energy Mater. 7, 1602121 (2017).

7. 7.

Guchhait, A. et al. Over 20% efficient CIGS-perovskite tandem solar cells. ACS Energy Lett. 2, 807–812 (2017).

8. 8.

Sahli, F. et al. Fully textured monolithic perovskite/silicon tandem solar cells with 25.2% power conversion efficiency. Nat. Mater. 17, 820–826 (2018).

9. 9.

Shockley, W. & Queisser, H. J. Detailed balance limit of efficiency of p–n junction solar cells. J. Appl. Phys. 32, 510–519 (1961).

10. 10.

Hailegnaw, B., Kirmayer, S., Edri, E., Hodes, G. & Cahen, D. Rain on methylammonium lead iodide based perovskites. Possible environmental effects of perovskite solar cells. J. Phys. Chem. Lett. 6, 1543–1547 (2015).

11. 11.

Tress, W., Correa Baena, J. P., Saliba, M., Abate, A. & Graetzel, M. Inverted current–voltage hysteresis in mixed perovskite solar cells. Polarization, energy barriers, and defect recombination. Adv. Energy Mater. 6, 1600396 (2016).

12. 12.

Leijtens, T. et al. Stability of metal halide perovskite solar cells. Adv. Energy Mater. 5, 1500963 (2015).

13. 13.

Arora, N. et al. Perovskite solar cells with CuSCN hole extraction layers yield stabilized efficiencies greater than 20%. Science 358, 768–771 (2017).

14. 14.

Li, Z. et al. Scalable fabrication of perovskite solar cells. Nat. Rev. Mater. 3, 18017 (2018).

15. 15.

Di Giacomo, F. et al. Up-scalable sheet-to-sheet production of high efficiency perovskite module and solar cells on 6-in. substrate using slot die coating. Sol. Energy Mater. Sol. Cells 181, 53–59 (2018).

16. 16.

Yang, M. et al. Perovskite ink with wide processing window for scalable high-efficiency solar cells. Nat. Energy 2, 17038 (2017).

17. 17.

Heo, J. H., Lee, M. H., Jang, M. H. & Im, S. H. Highly efficient CH3NH3PbI3−xClx mixed halide perovskite solar cells prepared by redissolution and crystal grain growth via spray coating. J. Mater. Chem. A 4, 17636–17642 (2016).

18. 18.

Zhou, Y., Game, O. S., Pang, S. & Padture, N. P. Microstructures of organometal trihalide perovskites for solar cells. Their evolution from solutions and characterization. J. Phys. Chem. Lett. 6, 4827–4839 (2015).

19. 19.

Cotella, G. et al. One-step deposition by slot-die coating of mixed lead halide perovskite for photovoltaic applications. Sol. Energy Mater. Sol. Cells 159, 362–369 (2017).

20. 20.

Schmidt, T. M., Larsen-Olsen, T. T., Carlé, J. E., Angmo, D. & Krebs, F. C. Upscaling of perovskite solar cells. Fully ambient roll processing of flexible perovskite solar cells with printed back electrodes. Adv. Energy Mater. 5, 1500569 (2015).

21. 21.

Das, S. et al. High-performance flexible perovskite solar cells by using a combination of ultrasonic spray-coating and low thermal budget photonic curing. ACS Photonics 2, 680–686 (2015).

22. 22.

Yang, Z. et al. High-performance fully printable perovskite solar cells via blade-coating technique under the ambient condition. Adv. Energy Mater. 5, 1500328 (2015).

23. 23.

Liu, J. et al. Improved crystallization of perovskite films by optimized solvent annealing for high efficiency solar cell. ACS Appl. Mater. Interfaces 7, 24008 (2015).

24. 24.

Yang, M. et al. Facile fabrication of large-grain CH3NH3PbI3−xBrx films for high-efficiency solar cells via CH3NH3Br-selective Ostwald ripening. Nat. Commun. 7, 12305 (2016).

25. 25.

Matsui, T., Seo, J.-Y., Saliba, M., Zakeeruddin, S. M. & Grätzel, M. Room-temperature formation of highly crystalline multication perovskites for efficient, low-cost solar cells. Adv. Mater. 29, 1606258 (2017).

26. 26.

Raga, S. R., Jiang, Y., Ono, L. K. & Qi, Y. Application of methylamine gas in fabricating organic–inorganic hybrid perovskite solar cells. Energy Technol. 5, 1750–1761 (2017).

27. 27.

Zhou, Y. & Padture, N. P. Gas-induced formation/transformation of organic–inorganic halide perovskites. ACS Energy Lett. 2, 2166–2176 (2017).

28. 28.

Zhao, T. et al. Design rules for the broad application of fast ( < 1 s) methylamine vapor based, hybrid perovskite post deposition treatments. RSC Adv. 6, 27475 (2016).

29. 29.

Singh, A., Chouhan, A. S. & Avasthi, S. Methylamine vapor exposure for improved morphology and stability of cesium-methylammonium lead halide. Preprint at https://arXiv.org/abs/1712.09068 (2017).

30. 30.

Chen, H. et al. A solvent- and vacuum-free route to large-area perovskite films for efficient solar modules. Nature 550, 92–95 (2017).

31. 31.

Zhou, Y., Yang, M., Pang, S., Zhu, K. & Padture, N. P. Exceptional morphology-preserving evolution of formamidinium lead triiodide perovskite thin films via organic-cation displacement. J. Am. Chem. Soc. 138, 5535–5538 (2016).

32. 32.

Ávila, J., Momblona, C., Boix, P. P., Sessolo, M. & Bolink, H. J. Vapor-deposited perovskites. The route to high-performance solar cell production?. Joule 1, 431–442 (2017).

33. 33.

Leyden, M. R. et al. High performance perovskite solar cells by hybrid chemical vapor deposition. J. Mater. Chem. A 2, 18742 (2014).

34. 34.

Remeika, M. & Qi, Y. Scalable solution coating of the absorber for perovskite solar cells. J. Energy Chem. 27, 1101–1110 (2018).

35. 35.

Zhou, Y. & Zhu, K. Perovskite solar cells shine in the “Valley of the Sun”. ACS Energy Lett. 1, 64–67 (2016).

36. 36.

Lee, J.-W. & Park, N.-G. Two-step deposition method for high-efficiency perovskite solar cells. MRS. Bull. 40, 654–659 (2015).

37. 37.

Bengtsson, L. A., Stegemann, H., Holmberg, B. & Füllbier, H. The structure of room temperature molten polyiodides. Mol. Phys. 73, 283–296 (1991).

38. 38.

Petrov, A. A. et al. A new formation strategy of hybrid perovskites via room temperature reactive polyiodide melts. Mater. Horiz. 4, 625–632 (2017).

39. 39.

Wang, S., Jiang, Y., Juarez-Perez, E. J., Ono, L. K. & Qi, Y. Accelerated degradation of methylammonium lead iodide perovskites induced by exposure to iodine vapour. Nat. Energy 2, 16195 (2016).

40. 40.

Rakita, Y., Gupta, S., Cahen, D. & Hodes, G. Metal to halide perovskite (HaP): an alternative route to HaP coating, directly from Pb(0) or Sn(0) films. Chem. Mater. 29, 8620–8629 (2017).

41. 41.

Zhang, W. et al. Ultrasmooth organic–inorganic perovskite thin-film formation and crystallization for efficient planar heterojunction solar cells. Nat. Commun. 6, 6142 (2015).

42. 42.

Shi, D. et al. Low trap-state density and long carrier diffusion in organolead trihalide perovskite single crystals. Science 347, 519–522 (2015).

43. 43.

Dong, Q. et al. Electron–hole diffusion lengths > 175 μm in solution-grown CH3NH3PbI3 single crystals. Science 347, 967–970 (2015).

44. 44.

Saliba, M. et al. Cesium-containing triple cation perovskite solar cells. Improved stability, reproducibility and high efficiency. Energy Environ. Sci. 9, 1989–1997 (2016).

45. 45.

Nie, W. et al. The effect of light soaking on the lattice change and the impact towards long-term photo-stability in Cs doped FA-MAPbI3 mixed cation perovskite alloys. In 3rd International Conference on Perovskite Solar Cells and Optoelectronics (PSCO-2017) A02.02 (2017).

46. 46.

Tang, Z. et al. Hysteresis-free perovskite solar cells made of potassium-doped organometal halide perovskite. Sci. Rep. 7, 12183 (2017).

47. 47.

Giordano, F. et al. Enhanced electronic properties in mesoporous TiO2 via lithium doping for high-efficiency perovskite solar cells. Nat. Commun. 7, 10379 (2016).

48. 48.

Correa Baena, J. P. et al. Highly efficient planar perovskite solar cells through band alignment engineering. Energy Environ. Sci. 8, 2928–2934 (2015).

49. 49.

Anaraki, E. H. et al. Low-temperature Nb-doped SnO2 electron-selective contact yields over 20% efficiency in planar perovskite solar cells. ACS Energy Lett. 3, 773–778 (2018).

50. 50.

Shin, S. S. et al. Colloidally prepared La-doped BaSnO3 electrodes for efficient, photostable perovskite solar cells. Science 356, 167–171 (2017).

## Acknowledgements

N.A.B., A.A.P., A.Yu.G., S.A.F., E.A.G. and A.B.T. acknowledge financial support from the Ministry of Education and Science of the Russian Federation, Project Number: RFMEFI60716X0147 and JSC “Krasnoyarskaya HPP”. I.T, S. Kazaoui, S.A. and T.U. acknowledge the support of the New Energy Development Organization of Japan. I.T. and S. Kazaoui thank Hiroshi Tomiyasu, Eisuke Ito (CEREBA) and Tetsuo Tsutsui (Kyushu University) for assistance. I.T. expresses gratitude to Anna Pavlova for her support.

## Author information

Authors

### Contributions

I.T. conceived the idea of the present work in discussion with A.B.T. Fabrication and characterization of perovskite solar cells and films by the RP-MAGIC method were conducted by I.T., S. Kazaoui and A.B.T. The Pb/MAI interface modification method was developed by S. Kosar through overlapped evaporation. S.A. and T.U. contributed to experiment preparation. N.A.B. and A.Yu.G. developed the spray-assisted RP-MAGIC method. S.A.F. and A.A.P. measured the melting temperatures of polyiodides. E.A.G. measured Raman spectra. I.T., S. Kazaoui, A.B.T., E.A.G., M.K. and M.G. performed scientific evaluation of the data. The manuscript was written by I.T., E.A.G., S. Kazaoui, S. Kosar, A.B.T. and M.G. The project was planned, directed and supervised by I.T. and A.B.T. All the authors discussed the results and commented on the manuscript.

### Corresponding author

Correspondence to Alexey B. Tarasov.

## Ethics declarations

### Competing interests

The authors declare no competing interests.

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

## Supplementary information

### Supporting information

Supplementary text and Supplementary Figs. 1–16

### Supplementary Video 1

Fabrication of large area MAPbI3 film on flexible PET/ITO substrate

### Supplementary Video 2

Conversion of the Pb/MAI bilayer into MAPbI3 with a crystallite of iodine as iodine source

## Rights and permissions

Reprints and Permissions

Turkevych, I., Kazaoui, S., Belich, N.A. et al. Strategic advantages of reactive polyiodide melts for scalable perovskite photovoltaics. Nature Nanotech 14, 57–63 (2019). https://doi.org/10.1038/s41565-018-0304-y

• Accepted:

• Published:

• Issue Date:

• ### Scalable fabrication and coating methods for perovskite solar cells and solar modules

• Nam-Gyu Park
•  & Kai Zhu

Nature Reviews Materials (2020)

• ### Performance improvement of perovskite solar cells via spiro-OMeTAD pre-crystallization

• Minghan Li
• , Yanyan Wang
• , Haoyuan Xu
• , Houcheng Zhang
• , Jing Zhang
• , Yuejin Zhu
•  & Ziyang Hu

Journal of Materials Science (2020)

• ### A chain-type diamine strategy towards strongly anisotropic triiodide of DMEDA·I6

• Li Yao
• , Peng Xu
• , Wanru Gao
• , Junze Li
• , Liang Gao
• , Guangda Niu
• , Dehui Li
• , Shiyou Chen
•  & Jiang Tang

Science China Materials (2020)