Perovskite solar cells (PSCs) have shown remarkable progress in efficiency as compared to CIGS and CdTe thin-film solar cells, however, instability of PSCs and toxicity of Pb remain issues for commercialization. While the quest to improve efficiency in CdTe, Cu(In,Ga)(S,Se)2 (CIGS), Cu(Zn,Sn)(S,Se)2 (CZTSSe) and related thin film solar cells has taken decades to reach the 10-23% level, organic-inorganic perovskites (e.g., CH3NH3PbI3-xClx) have emerged with a relatively quick demonstration of efficiencies of 22.1% and simple low-temperature, low-cost processing. The use of a tandem structure could realistically improve efficiency of a CIGS device (22.3%) to beyond 30%. Perovskite tandem solar cells have shown efficiencies above 25% with silicon and 17.8% with CIGS bottom cells respectively. Despite these attractive attributes, there are areas that need to be addressed on the fundamental materials level such as materials stability and replacement of Pb with a non-toxic and sustainable alternative. CsSnI3 is a highly desirable substitute for Pb-based perovskites, but poor stability of the material has prevented fabrication of devices that can withstand sustained operation or even processing under ambient air conditions. The inorganic compound Cs2SnI6 has received recent attention as an alternative to Sn-based halide perovskites for photovoltaic device applications. In comparison to Sn- and Pb-based halide perovskites, Cs2SnI6 has been shown to feature enhanced stability in ambient environments due to the stable oxidation state of Sn (Sn4+ in Cs2SnI6 as compared to Sn2+ in CsSnI3). The structure of Cs2SnI6 has been described as a defect variant of AMX3 perovskites, with half of the Sn atoms removed resulting in shorter Sn-I bond length and hence improved stability.
In this paper, we demonstrate a ~1.6 eV direct bandgap SnF2 doped Cs2SnI6 films processed via solution processing in air. The films are stable when annealed in dark upto 1000 hours at 100 C. Cs2SnI6 compound has been reported to display intrinsic n-type conductivity (with carrier concentrations of 10^14 cm-3 and 5 x 10^16 cm-3), and it has been shown that it can be doped p-type with SnI2 (with carrier concentrations of 10^14 cm-3), demonstrating the ambipolar nature of this material. Here we demonstrate a p-type Cs2SnI6 film with carrier concentration of 5 x 10^14 cm-3 and mobility of about 18 cm2/V-sec, which is comparable to CH3NH3PbI3. Effect of doping at different concentration of SnF2 will be discussed on carrier concentration and hole mobility. SCAPS calculation results for improving device efficiency will also be discussed. We have observed a strong effect of solvent (GBL, DMF or DMSO) on the resulting Cs2SnI6 phase, which will also be discussed here. The serendipitous discovery of this simple solution process sheds new light on possible properties of Cs2SnI6 which have not yet been reported experimentally.