Interface Passivation in Perovskite Photovoltaics: Formation, Dimensionality and Photophysics

Achieving net zero carbon emissions requires a 30-fold increase in solar power within 30 years. To reduce costs, cheaper alternatives to Si photovoltaics are being explored. Metal halide perovskites are defect-tolerant semiconductors, meaning low-cost manufacturing methods can produce thin films with outstanding optoelectronic properties. However, perovskite cells suffer from relatively poor stability due to their soft crystal structure and high mobile ion densities. As interfaces have the highest defect densities, successful methods to improve stability and efficiency focus on perovskite film surfaces. In this thesis I present methods to passivate perovskite interfaces for improved performance and operating lifetime.I begin by exploring the formation of 2D/3D perovskite heterostructures. Converting the top surface of a perovskite film into 2D perovskite creates a barrier to ion movement and passivates surface defects. Despite their use in high efficiency devices, the formation mechanism of the 2D layer was unknown. I found three formation routes that depend on solvent and perovskite composition. By using non-polar solvents, I was able to slow down the 2D conversion reaction to produce more uniform 2D layers improving power conversion efficiency (PCE) from 20% to 21.9%. Next, I demonstrate that modifying the dimensionality of the 2D layer alters its energetic alignment. I present two methods to tune dimensionality and use drift-diffusion modeling to explain resulting photovoltaic performance. Previous studies have shown that 2D/3D heterostructures have the correct band alignment to improve only the hole contact in n-i-p configured cells. Herein I demonstrate dimensional tuning to produce p-i-n perovskite solar cells with 23.9% PCE (certified quasi-steady state, NREL). When subjected to ISOS-L3 accelerated ageing, these devices retain 92% of initial PCE after 500 h. Finally, I describe a method to alter the surface energetics of a wider bandgap perovskite absorber (1.8 eV). In this case, no 2D layer is formed. Instead, the 1,3–propanediammonium surface treatment removes structural heterogeneity and improves band alignment with a C60 (fullerene) electron transporting layer. When built into a monolithic tandem, PCE is 27.4%, higher than the best single junction Si cell. These tandems retained more than 86% of their initial PCE after 500 h of operation.