On the Equilibrium Electrostatic Potential and Light-Induced Charge Redistribution in Halide Perovskite Structures

Lead halide perovskites (LHP) are semiconductor materials which are employed as not-intentionally-doped light absorbers inserted between two selective carrier transport layers (SCTL), realizing a p-i-n or n-i-p heterojunction. Despite the recent achievements in power conversion efficiency of LHP solar cells, there is still a large debate about the role and concentration of defects. Being polycrystalline materials, LHPs can contain defects both in the bulk and in the grain boundaries. Therefore, the combination of surface characterization techniques and modelling could unravel some of those defects’ key properties, such as their concentration, type, energy level, and capture rates.<br/>In this work, we present a characterization and modelling study conducted to gain insight into perovskite doping levels and defect concentrations. We produced a lateral heterojunction (LHJ) device by taking a glass substrate coated with fluorine-doped tin oxide, a transparent conductive oxide (TCO). We selectively deposited TiO_x and NiO_x, and we opened a long channel (&gt;60 μm) between the two materials employing laser scribing. On top, we deposited a methylammonium lead iodide perovskite by spin coating.<br/>The free perovskite surface was locally probed using X-ray photoemission spectroscopy (XPS). This allowed us to acquire the electrostatic potential profile (φ(x)) along the channel by measuring the shift of the I 3d_5/2 core level peak energy (<i>I</i>_3d5/2) with respect to the Fermi level (<i>E</i>_F): qφ(x)=<i>E</i>_F-<i>I</i>_3d5/2(x).<br/>We obtain a gradual monotonous decrease of the potential in the channel, from the top of TiO_x to that of NiO_x. 2D drift-diffusion simulations were performed to investigate the doping and defects in the layer. Our study suggests that the effective doping level in halide perovskites should be very low, below 10¹³ cm^-3. Along the vertical direction, this doping level is not enough to screen the electric field of the buried heterojunction with a transport layer. The perovskite work function is therefore impacted by the SCTL work function.<br/>We also performed Kelvin probe force microscopy (KPFM) experiments on perovskite-SCTL-TCO structures, measuring surface photovoltages (SPVs) of hundreds of mVs under continuous illumination. Drift diffusion simulations show that an intrinsic perovskite – highly doped SCTL structure develops only few mV SPV under continuous illumination, due to the presence of a very low built-in voltage across the buried heterojunction. Therefore, the large SPVs found in experiments could be linked to deviations from these “ideal” characteristics.<br/>We showed that the presence of band offsets at the buried heterojunctions (perovskite-SCTL or SCTL-TCO), could considerably increase charge separation between layers. Consequently, the amplitude of the SPV would increase, which can be explained by thermionic emission considerations.<br/>In our simulations, we also considered the presence of traps in the SCTL: given the selective nature of the layer, the carriers injected by the perovskite could be trapped in the SCTL. This process could increase charge separation and SPV.<br/>To confirm the previous experimental results and to further study the influence of contact materials on the perovskite electronic properties, a new LHJ device is under development. This device will be produced employing photolithography techniques and will feature different channel lengths. Photolithography will allow to reach micrometric resolution on the channel width while maintaining low roughness, which will make possible to perform KPFM measurements of the surface potential across the channels, both in dark and under illumination. New contact materials will be employed, like indium tin oxide or aluminum doped zinc oxide, and new SCTLs will be tested, like SnO_x or self assembled hole transport layers.