Chloride in lead-chloride derived organo-metal halides for perov- skite-absorber solar cells Eva L. Unger, ‡a† Andrea R. Bowring, ‡a Christopher J. Tassone, b Vanessa L. Pool, b Aryeh Gold- Parker, b,c Rongrong Cheacharoen, a Kevin H. Stone, b Eric T. Hoke, a Michael F. Toney b and Mi- chael D. McGehee a* a Department of Materials Science and Engineering, Stanford University, Stanford, California 94305, United States b Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, Menlo Park, California 94025, United States c Department of Chemistry, Stanford University, Stanford, California 94305, United States KEYWORDS Perovskite, Methylammonium lead iodide, Solar cells, Chloride, Crystallization ABSTRACT: Organo-metal halide perovskites are an intriguing class of materials that have recently been explored for their potential in solar energy conversion. Within a very short period of intensive research, highly efficient solar cell devices have been demonstrated. One of the heavily debated questions in this new field of research concerns the role of chlorine in solution-processed samples utilizing lead chloride and three equivalents of methylammonium iodide to prepare the perovskite samples. We utilized a combination of X-ray photoelectron spectroscopy, X-ray fluorescence and X-ray diffraction to probe the amount of chlorine in samples before and during annealing. As-deposited samples, before annealing, consist of a crystalline precursor phase containing excess methylammonium and halide. We used in- situ techniques to study the crystallization of MAPbI 3 from this crystalline precursor phase. Excess methylammonium and chloride evaporate during annealing forming highly crystalline MAPbI 3 . However, even after prolonged annealing times chlorine can be detected in the films in X-ray fluorescence measurements. Introduction Perovskite-absorber solar cells have become a research area of high expectations for their potential in revolu- tionizing solar energy conversion technology. Seminal work utilizing hybrid organic-inorganic absorber mate- rials as light-harvesters in liquid-electrolyte dye- sensitized solar cells demonstrated promising efficiency metrics but low stability due to the rapid dissolution of the perovskite absorber in the liquid electrolyte envi- ronment. 1,2 After replacing the liquid electrolyte with a solid hole transporting material, efficiencies of solar cell devices incorporating the perovskite-absorber methyl- ammonium lead iodide (MAPbI 3 ) have undergone an unprecedented climb in device efficiency over the past two years. 3–11 High solar energy conversion efficiencies have been achieved both for device-architectures de- rived from solid-state dye-sensitized solar cells as well as devices utilizing an inert metal oxide as a scaffold and thin film devices. In the first reports, the organo-metal halide was de- posited in one step from a solution containing lead io- dide (PbI 2 ) and methylammonium iodide (MAI) in equimolar ratio dissolved in gamma-butyrolactone. 1,2,4 Lee et al. 3 reported a different deposition approach using lead chloride (PbCl 2 ) as the source for lead and a three- fold excess of methylammonium iodide. Compared to the equimolar deposition route, 1,2,4 this deposition method requires annealing at 100°C for 45 minutes until the perovskite absorber is fully formed. While early re- ports suggested a composition of MAPbClI 2 the exact amount and role of chlorine in PbCl 2 -derived organo- metal-halide perovskites has been debated 12–15 and the stoichiometry of this material is frequently stated as MAPbCl x I 3-x en lieu of a more defined stoichiometry. 5,8,15– 17 While the bandgap of MAPbCl x I 3-x is nearly identical to MAPbI 3 , 3,4,12,15 the chlorine-derived compounds have been shown to exhibit superior charge carrier diffusion lengths. 17,18 Infiltrated into meso-porous titania, MAPbCl x I 3-x had a considerably lower series resistance compared to pure MAPbI 3 . 12 The differences in the op- toelectronic properties 19 could thus be due to chlorine doping, 3,12 passivation of trap-states or differences in sample crystallinity. 14,20–22 Unlike the mixed bromide- iodide methylammonium lead halides, 7 the chloride- iodide mixed halide perovskites exhibit low miscibility. 12 MAPbCl x I 3-x thin film samples are often found to be highly oriented with the (110) direction perpendicular to the substrate. In this case, no scattering information is collected from other crystallographic directions in con- ventional θ-2θ-X-ray diffraction (XRD) measurements, preventing the full determination of the crystal struc- ture. Early work speculated therefore that chloride is incorporated into the perovskite structure along the (001)-direction. 15 Powder-XRD analysis of scraped-off PbCl 2 -derived thin films exhibited lattice parameters suggesting a stoichiometry close to pure MAPbI 3 with a