Strain regulation retards natural operation decay of perovskite solar cells

Abstract

Perovskite solar cells (pero-SCs) have undergone rapid development in the past decade. However, there is still a lack of systematic studies investigating whether the empirical rules of working lifetime assessment used for silicon solar cells can be applied to pero-SCs. It is believed that pero-SCs show enhanced stability under day/night cycling owing to the reported self-healing effect in the dark^1,2. Here we find that the degradation of highly efficient FAPbI₃ pero-SCs is much faster under a natural day/night cycling mode, bringing into question the widely accepted approach to estimate the operational lifetime of pero-SCs based on continuous-mode testing. We reveal the key factor to be the lattice strain caused by thermal expansion and shrinking of the perovskite during operation, an effect that gradually relaxes under the continuous-illumination mode but cycles synchronously under the cycling mode^3,4. The periodic lattice strain under the cycling mode results in deep trap accumulation and chemical degradation during operation, decreasing the ion-migration potential and hence the device lifetime⁵. We introduce phenylselenenyl chloride to regulate the perovskite lattice strain during day/night cycling, achieving a certified efficiency of 26.3 per cent and a 10-fold improvement in the time required to reach 80% of peak efficiency (T₈₀) under the cycling mode after the modification.

Fig. 1. Faster PCE decay and cycled lattice strain in the cycling mode.

a, Stability of the pero-SCs based on FAPbI₃ working in the continuous and cycling modes. The purple line represents the average value for each cycling mode with illumination time. b, Stability of the pero-SCs based on FAPbI₃ working in the continuous and cycling modes. The temperature of the device was fixed at room temperature (RT; about 25 °C). c, Stability of the pero-SCs based on FAPbI₃ working in the continuous thermal and cycled thermal (12 h at room temperature and 12 h at about 55 °C) ageing modes. Error bars represent the standard deviation of five devices for each condition. d,e, Integrated profiles (out of plane) obtained from in situ GIWAXS maps for the FAPbI₃ perovskite film deposited on a silicon wafer substrate. The film was illuminated for 0 to 90 min and measured at 30-min intervals for the two cycles (d, the first cycle; e, the second cycle), and the recovery spectra were obtained from the film kept in dark for 30 min. f, Evolution of the unit cell volume and the peak broadening parameters for the FAPbI₃ samples during in situ GIWAXS measurements. r.u., relative units. Source Data

Fig. 2. Defect evolution, ion-migration dynamics and stress evolution in the cycling mode.

a, The tDOS spectra of the pero-SCs before and after ageing under the continuous mode (168 h) and the cycling mode (14 cycles). E_ω represents the energetic distribution, as mentioned in Supplementary Note 3. b, Spatial distribution of the trap densities of trap band II in the pero-SCs before and after ageing under the continuous mode (168 h) and the cycling mode (14 cycles). SnO₂ and spiro-OMeTAD in the graphs indicate the locations that are close to the SnO₂ or spiro-OMeTAD layers of the device. c, E_a of the FAPbI₃ device before and after ageing in the continuous mode and the cycling mode. d,e, Illustration of the degradation mechanism for the pero-SCs in the continuous mode and the cycling mode. Perovskite film stress was calculated from curvature measurements, as mentioned in Supplementary Note 2. Source Data

Fig. 3. Mitigating cycled lattice strain through Ph-Se-Cl modification.

a, X-ray diffraction of the wet perovskite films (before annealing) with different amounts of Ph-Se-Cl. b, The peak ratio of α-FAPbI₃/PbI₂ and intermediate phase/PbI₂. c, High-resolution transmission electron microscopy images of the pero-Ph-Se-Cl film. Parameters d represents the interplanar spacing. d, Structural refinement (La Bail method) of the integrated GIWAXS profile recorded from the pero-Ph-Se-Cl films. e, Normalized (Nor.) lattice parameters and calculated spontaneous strain of FAPbI₃ and the pero-Ph-Se-Cl samples. Parameters ε_tet and ε_orth respectively represent the degenerate tetragonal and orthorhombic symmetry-adapted strains which emerge during the phase transitions. Parameters a, b, and c are the normalized lattice parameters of the perovskite; a₀ is estimated by taking the cube root of the normalized unit cell volume. f,g, Integrated profiles obtained from in situ GIWAXS maps of the pero-Ph-Se-Cl films (f, the first cycle; g, the second cycle). h, Evolution of the unit cell volume and the peak broadening parameters for the pero-Ph-Se-Cl samples during the in situ GIWAXS measurements. Source Data

Fig. 4. Stability of the pero-Ph-Se-Cl-based devices.

a, Current–voltage curves of the pero-SCs. b, Stability of the pero-SCs based on FAPbI₃/PEAI/spiro-OMeTAD in the cycling mode. c, Stability of the pero-SCs based on FAPbI₃/BDT-DPA-F in the cycling working. The temperature of the device fluctuated from about 85 °C under illumination to room temperature in the dark. d, Stability statistical graphs of the pero-SCs based on FAPbI₃, FA_0.92MA_0.08PbI₃ and Cs_0.05FA_0.7MA_0.25PbI_2.6Br_0.4 working in the continuous mode (108 h) and the cycling mode (9 day/night cycles, illumination 108 h). Error bars represent the standard deviation of 10 devices for each condition. Source Data

Extended Data Fig. 1. Faster PCE decay of pero-SCs based on FAPbI₃ in the day/night cycling working mode.

a-c, PCE tracking of the pero-SCs based on FAPbI₃/carbon electrode working in the continuous-illumination and day/night cycling modes (a, day period: ~55 °C; b, day period: ~65 °C; c, day period: ~85 °C). Source Data

Extended Data Fig. 2. Faster PCE decay of pero-SCs based on other perovskite active layers in the day/night cycling working mode.

a,b, PCE tracking of the pero-SCs based on a, FA_0.92MA_0.08PbI₃ and b, Cs_0.05FA_0.7MA_0.25PbI_2.6Br_0.4 working in the continuous-illumination and day/night cycling modes. Source Data

Extended Data Fig. 3. The orthorhombic phase for FAPbI₃ at RT.

a,b, Temperature-dependent XRD patterns (cooling process) of perovskite films based on FAPbI₃ (focused on (002)/(110) and (004)/(220) diffractions, respectively). c, Comparison of structural refinements made using cubic (α-phase), tetragonal (β-phase) and orthorhombic (γ-phase) perovskite structures. The goodness of fit values (χ²) is inset, confirming that the best fit is made with a γ-phase structure at RT. Source Data

Extended Data Fig. 4. The performance of pero-SCs with different chalcogenides.

a, SEM images of perovskite films based on different chalcogenides. The scale bar is 1 µm. b-d, Photovoltaic parameters for the devices with different concentration of different chalcogenides under AM1.5 G illumination (10 devices for each type). Source Data

Extended Data Fig. 5. The formation of PhSe-plumbate.

a, XRD pattern of pure PbI₂ powder, PbI₂-DMF single crystal and PhSe-plumbate powders. b, The corresponding pictures. c, ¹H NMR spectrum of PhSe-plumbate. d, ⁷⁷Se NMR spectrum of PhSe-plumbate. Source Data

Extended Data Fig. 6. The pseudo-cubic phase for pero-Ph-Se-Cl.

Temperature-dependent XRD patterns (cooling process) of the pero-Ph-Se-Cl film. Source Data

Extended Data Fig. 7. Thermal stability of the pero-Ph-Se-Cl-based devices.

(Device structure: FTO/SnO₂/FAPbI₃/BDT-DPA-F/Au). a, PCE tracking of the pero-SCs without and with Ph-Se-Cl modification under 85 °C in a N₂-filled glovebox. Data from six cells are collected and presented as mean values  ±  standard error of the mean. b, PCE tracking of the pero-SCs without and with Ph-Se-Cl modification under the cycled thermal (12 h RT and 12 h 85 °C) working modes according to the ISOS-T-1 suggested protocol. Source Data

Extended Data Fig. 8. The narrow PCE-degradation gap between the two working modes.

a-c, PCE tracking of the pero-SCs with Ph-Se-Cl modification based on a, FAPbI₃, b, FA_0.92MA_0.08PbI₃ and c, Cs_0.05FA_0.7MA_0.25PbI_2.6Br_0.4 three different active layer compositions under the continuous-illumination and day/night cycling working modes. Source Data