Development of an all-vapor process using two-step close space vapor transport for the deposition of perovskite solar cells
Date
2022
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
Lead halide perovskites are a promising next generation photovoltaic material with impressive laboratory scale results. To achieve industrial relevance, scalable deposition processes must be developed. Here, an all-vapor process for the deposition of perovskite solar cells using two-step close space vapor transport is developed. ☐ The effect of the electron transport layer (ETL) on the growth of PbI2 thin films, reaction into methylammonium lead iodide (MAPbI3), and impact on photovoltaic performance is investigated, including the effect on hysteresis and air-stability. The J-V performance for optimized fabrication conditions on each ETL (SnO2, C60, and CdS) was similar but the ETL affected hysteresis and air stability behavior. C60 ETLs reduced the hysteresis in the solar cells, in agreement with previous literature. However, C60 is found to cause loss of fill factor (FF) and the formation of an s-shaped current voltage (J-V) curve within 15 minutes due to interactions with air that reduce its conductivity and hinder charge transport in the solar cell. The design and use of passivating layers on SnO2 ETLs can result in a better combination of hysteresis and air-stability. ☐ The effect of deposition parameters on the growth rate and morphology of PbI2 thin films is briefly investigated over a wide parameter space to enable the deposition of reproducible, compact films that are useful for solar cell applications. The effect of reaction parameters on the conversion of PbI2 to MAPbI3 is also investigated, where temperature and time can lead to significant differences in the reaction rate. The amount of residual PbI2 is found to be a critical factor for solar cell performance. Solar cells where all PbI2 was consumed exhibit reduced performance, likely due to reactions between the ETLs and the perovskite or excess MAI introduced to complete the reaction. Possible degradation mechanisms are identified through comparison to previous reports. Solar cells fabricated at optimal conditions without post-treatments show 12-14% photovoltaic conversion efficiency (PCE), primarily limited by open circuit voltage (Voc) and FF. Therefore, analysis to identify the limiting factors of these solar cells was carried out, including comparison to record solar cell J-V parameters. ☐ FF losses come from a combination of series resistance and charge collection losses, as well as nonradiative recombination. Series resistance losses can be quantified using J-V analysis and collection losses can be quantified using JscVoc measurements. Voc losses come from nonradiative recombination losses, which are investigated by drift-diffusion modeling, drive level capacitance profiling (DLCP), and time-resolved photoluminescence (TRPL). Based on the hypothesis that the two-step reaction causes an excess of methylammonium iodide (MAI) in the perovskite that leads to high nonradiative recombination via iodine interstitial defects at the reaction interface and bulk of the film, a methylamine gas recrystallization treatment was applied to cause liquefaction and mixing of the film, which improved the photovoltaic performance in a manner consistent with the expected effect from drift-diffusion simulations. ☐ Copper Phthalocyanine (CuPC) is developed as a vapor-processable hole transport layer (HTL). As-deposited CuPC films in an inert environment led to poor performance attributed to high resistivity causing low FF and s-shaped J-V curve characteristics due to poor carrier extraction. In-situ or ex-situ exposure to O2 combined with a MoOx hole extraction layer improves the conductivity of the CuPC films and enables 12-14% PCE, well-behaved solar cells with FF up to 72%, presumably caused by O2 doping of CuPC. Solar cells with spin coated spiro-OMeTAD films, made for comparison, show similar PCE but higher Voc. Drift-diffusion simulations are used to hypothesize that the difference in Voc is mainly driven by the valence band position and doping densities of the HTLs, and their effect on surface recombination and carrier extraction, respectively. ☐ This work has established an all-vapor process to reproducibly make perovskite solar cells with PCE up to 15.8%. The observations herein can be used to develop additional studies to further improve performance, as well as design, scale up, and troubleshoot similar processes such as vapor transport deposition with a goal of manufacturing perovskite solar cells on an industrial scale.
Description
Keywords
All-vapor processing, Charge transport layer, Close space vapor transport, Lead halide perovskites, Thin film solar cells