Lead Halide Residue as a Source of Light- Induced Reversible Defects in Hybrid Perovskite Layers and Solar Cells Jakub Holovský ,* ,†,‡ Amalraj Peter Amalathas, † Lucie Landova ́ , †,‡ Branislav Dzurň a ́ k, † Brianna Conrad, † Martin Ledinský , ‡ Zdeň ka Ha ́ jkova ́ , ‡ Ognen Pop-Georgievski, § Jan Svoboda, § Terry Chien-Jen Yang, ∥ and Quentin Jeangros ∥ † Centre for Advanced Photovoltaics, Faculty of Electrical Engineering, Czech Technical University in Prague, Technická 2, 16627 Prague, Czech Republic ‡ Institute of Physics, AS CR v. v. i., Prague, Cukrovarnická 10, 16200 Prague, Czech Republic § Department of Chemistry and Physics of Surfaces and Biointerfaces, Institute of Macromolecular Chemistry, Czech Academy of Sciences, Heyrovsky sq. 2, 16206 Prague, Czech Republic ∥ Institute of Microengineering (IMT), Photovoltaic and Thin-Film Electronics Laboratory, E ́ cole Polytechnique Fé dé rale de Lausanne (EPFL), Rue De La Maladiè re 71b, 2002 Neuchâ tel, Switzerland * S Supporting Information ABSTRACT: Advanced characterization methods avoiding transient effects in combination with solar cell performance monitoring reveal details of reversible light-induced perovskite degradation under vacuum. A clear signature of related deep defects in at least the 1 ppm range is observed by low absorptance photocurrent spectroscopy. An efficiency drop, together with deep defects, appears after minutes-long blue illumination and disappears after 1 h or more in the dark. Systematic comparison of perovskite materials prepared by different methods indicates that this behavior is caused by the lead halide residual phase inherently present in material prepared by the two-step method. X-ray photoelectron spectros- copy confirms that lead halide when illuminated decomposes into metallic lead and mobile iodine, which diffuses into the perovskite phase, likely producing interstitial defects. Single-step preparation, as well as preventing lead halide illumination, eliminates this effect. H ybrid metal halide perovskites have recently proven their prophesied capabilities. First, they have yielded record efficiencies (25.2%) among polycrystalline thin film technologies. 1 Second, they have been used with crystalline silicon cells to overcome the latter’s single-junction limit by forming a highly efficient (28%) tandem device. 1 We have previously shown that the sharp band edge is an important prerequisite for such success. 2,3 However, a low value of absorption below the band edge indicating a low concentration of deep defects is an equally important prerequisite also deserving of careful investigation that must respect material differences. Because the advantage of perov- skite materials is easy preparation, a great variety of materials and fabrication techniques are in use, which is vital for such rapid device development. For the best single-junction solar cells, 4 single-step solution-processing is typically used, 5 and in the best nonindustrial tandem device, 6 the two-step process is used, 7 but these processing techniques are considerably different. This should be respected when focusing on defects and especially the defects related to instability. Stabilized efficiencies (19%), 8 are much lower than the previously mentioned records. Stability issues, which stem from material’s susceptibility to perturbations by illumination, exposure to humidity, various gases, or voltage bias, represent currently the biggest challenge for future commercialization. 9 To understand the stability of perovskites, research not only has focused on mixed-halide triple-cation materials 10 but has recently also returned to prototype pure lead or bromide halide perovskites, 11−13 which are generally less stable but simpler to understand. Sophisticated models of trap states calculated by first-principles, 11,14−16 charging, 12,17,18 structural rearrange- ment, 19 redistribution, 20 creation of hydrate, 21 and eventual decomposition into the metal halide and other products have Received: September 23, 2019 Accepted: November 13, 2019 Published: November 13, 2019 Letter http://pubs.acs.org/journal/aelccp Cite This: ACS Energy Lett. 2019, 4, 3011-3017 © XXXX American Chemical Society 3011 DOI: 10.1021/acsenergylett.9b02080 ACS Energy Lett. 2019, 4, 3011−3017 Downloaded via CSIRO on December 1, 2019 at 09:56:04 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.