Reversible Hydration of CH3NH3PbI3 in Films, Single Crystals, and Solar Cells

TLDR: In this paper, it was shown that hydrated crystal phases are formed when methylammonium lead iodide perovskite (MAPI) is exposed to water vapor at room temperature and these phase changes are fully reversed when the material is subsequently dried.

Abstract: Solar cells composed of methylammonium lead iodide perovskite (MAPI) are notorious for their sensitivity to moisture. We show that (i) hydrated crystal phases are formed when MAPI is exposed to water vapor at room temperature and (ii) these phase changes are fully reversed when the material is subsequently dried. The reversible formation of CH3NH3PbI3·H2O followed by (CH3NH3)4PbI6·2H2O (upon long exposure times) was observed using time-resolved XRD and ellipsometry of thin films prepared using “solvent engineering”, single crystals, and state-of-the-art solar cells. In contrast to water vapor, the presence of liquid water results in the irreversible decomposition of MAPI to form PbI2. MAPI changes from dark brown to transparent on hydration; the precise optical constants of CH3NH3PbI3·H2O formed on single crystals were determined, with a bandgap at 3.1 eV. Using the single-crystal optical constants and thin-film ellipsometry measurements, the time-dependent changes to MAPI films exposed to moisture were m...

Figures

Table 1. Critical point energies obtained for the best fit of MAPI and its hydrate. The complete model of the hydrate is given in Supplementary Information (Table S2).

Figure 5. Photographs of thin films of MAPI deposited on glass at different hydration times showing discoloration and AFM measurements (3D and 2D representations).

Figure 3. Time-resolved XRD patterns of polycrystalline MAPI: (a) hydration of a MAPI tin thin film and (b) dehydration of directly synthesized hydrate needle-like crystals.

Figure 4. (a) The Bruggeman-type degradation mechanism used to model the ellipsometry

Figure 6. (a) and (b) Current-voltage characteristics showing full recovery of a planar heterojunction MAPI solar cell upon exposure to moisture; scanned from reverse to forward

Figure 2. (a) Identification of the composition of the hydrate species grown on MAPI single

Frequently asked questions

Q1. What have the authors contributed in "The reversible hydration of ch3nh3pbi3 in films, single crystals and solar cells" ?

The authors show that hydrated crystal phases are formed when MAPI is exposed to water vapour at room temperature and that these phase changes are fully reversed when the material is subsequently dried. The reversible formation of CH3NH3PbI3•H2O followed by ( CH3NH3 ) 4PbI6•2H2O ( upon long exposure times ) was observed using time resolved XRD and ellipsometry of thin films prepared using ‘ solvent engineering ’, single crystals, and state of the art solar cells. Hysteresis in the current-voltage characteristics was significantly increased after this dehydration, which may be related to changes in the defect density and morphology of MAPI following recrystallization from the hydrate. The results suggest that the mono-hydrate phase forms independently of the depth in the film suggesting rapid transport of water molecules along grain boundaries. Vapour phase hydration of an unencapsulated solar cell ( initially Jsc ≈ 19 mA cm -2 and Voc ≈ 1. 05 V at 1 sun ) resulted in more than a 90 % drop in short circuit photocurrent and around 200 mV loss in open circuit potential, but these losses were fully reversed after the cell was exposed to dry nitrogen for 6 hours. Based on their observations the authors suggest that irreversible decomposition of MAPI in the presence of water vapour only occurs significantly once a grain has been fully converted to the hydrate phase. 

Q2. How long did the substrate be annealed?

After 30 s total spinning time, the substrate was immediately annealed at 100 °C for 10 min to evaporate residual solvents and to further promote crystallization. 

Q3. What is the effect of water vapour on the film?

Since the presence of water vapour appears to catalyse dynamic recrystalisation within the film between the hydrated and pure crystalline phases, this may lead to higher quality films under the optimised processing conditions as long as water is subsequently completely removed by thermal annealing. 

Q4. How long did the substrates be heated?

To complete the transformation of TiOx into the anatase phase, the coated substrates were gradually heated to 500 °C (ramp 8 °C/min) and sintered for 45 min in air. 

Q5. Why is the water released as the reaction goes from the monohydrate to the dihydrate?

The release of water as the reaction goes from the monohydrate to the dihydrate suggests partial self-sustainability of the conversion process as the water molecules released can be reused to convert remaining MAPI into the monohydrate. 

Q6. What is the biggest barrier to commercializing MAPI?

3Ensuring the stability of MAPI photovoltaics under operational conditions is one of the biggest barriers to commercializing the technology. 

Q7. What was the result of the hydration of the MAPI crystals?

The product obtained was metastable and turned into greyish polycrystalline MAPI by spontaneous loss of its crystalline water under ambient conditions. 

Q8. What was the effect of the patterned mask on the evaporation of gold?

a 40 nm thick gold layer was deposited by thermal evaporation through a patterned shadow mask under high vacuum conditions (4 × 10-6 mbar) to form the counter electrode. 

Q9. How many minutes did the spectra of a crystalline MAPI film be recorded?

Ellipsometry spectra were recorded every 10 or 20minutes during the conversion of a crystalline MAPI film to CH3NH3PbI3•H2O by exposure to air with RH 80 % over a period of 100 minutes. 

Q10. What are the three fit parameters in the model?

The three fit parameters in this model are: the thickness of the solid thin film, the thickness of the roughness layer on top of the film, and the relative ratio of MAPI and its hydrate in the mixture forming the film and roughness layer. 

Q11. What is the effect of the XRD on the crystalline MAPI?

The reduction of this peak intensity is accompanied by a decrease of the peak intensities associated with MAPI monohydrates and a concomitant increase in the intensity of the crystalline MAPI peaks. 

Q12. How can the hydrate content of the solid layer be estimated?

14,15 Using the fitted hydrate content as a function of exposure time (Figure 3 in the main text), values of the increase of the solid layer thickness due to its expansion on partial hydration can be estimated. 

Q13. How long did the substrates sit on the hotplate?

The substrates were coated with the TiOx sol-gel solution by spin-coating dynamically at 2000 rpm for 45 s and then quickly placed on a hotplate at 150 °C for 10 min. 

Q14. What is the simplest way to describe the surface roughness of a composite?

Surface roughness between two layers is also commonly described by an EMA layer consisting of a composite of top and bottom material in equal proportion. 

Q15. Why is dehydration faster than hydration at room temperature?

The authors observed that for crystallites of similar sizes, dehydration appears to be a faster process that hydration at room temperature. 

Q16. How does the structure of MAPI differ from other monohydrates?

Although not yet widely appreciated, MAPI clearly shows a propensity to form new solvated crystal structures at room temperature by incorporating small polar molecules. 

Q17. How much light is lost after exposure to moisture?

It is apparent for both scan directions that there is almost an order of magnitude drop in the photocurrent and around a 200 mV loss in photovoltage after the device was exposed to moisture for 3 hours. 

Q18. What is the XRD pattern of the MAPI needle shaped crystals?

Dihydrate crystals are obtained –together with the monohydrate species– in directly synthesized hydrated MAPI needle shaped crystals prepared from solution (methods section, photographs in Figure S4, characteristic reflection at 2θ = 11.39° in XRD pattern in Figure 2a).