Crystalline silicon (c-Si) solar cells have been dominating the photovoltaic market for decades. However, with the current record efficiency of 25.6 %, they almost reached their practical efficiency limit. It will therefore not be possible to significantly improve cell performance by simple process optimization. One of the most promising approaches to increase cell performance at reasonable costs lies in combining a c-Si cell with a high-bandgap cell to form a tandem device, allowing for reduced thermalization losses and a better utilization of the incident sunlight’s spectrum.

Organic-inorganic metal halide perovskite materials are particularly suited for this application due to their properties, such as a wide (and tunable) band gap, a steep absorption edge, and low sub-gap absorption. Furthermore, they can be produced with relatively cheap wet-chemical methods.

In this poster, we present recent advances we made in developing lead halide perovskite top cells, which are intended for monolithic perovskite/c-Si tandems. In this tandem configuration, the perovskite cell is directly fabricated on top of the c-Si cell. This reduces parasitic absorption losses between the two cells and needs fewer manufacturing steps, thus reducing cost. However, it imposes compromises on both sub-cells, e. g. a non-textured front surface for the c-Si cell and the limitation of the perovskite cell fabrication to low temperatures if temperature-sensitive c-Si cells, such as silicon heterojunction cells, are used.

To optimise perovskite cells for tandem applications, we developed a low-temperature planar perovskite cell based on a fullerene electron transport layer, the sequential evaporation of PbI₂ and spin-coating of methylammonium iodide (MAI), and a spiro-OMeTAD hole transport layer. By varying the MAI solvent concentration and using additives, we strongly improved the uniformity of the perovskite absorber layer, while reducing its surface roughness to an R_q of less than 10 nm. In addition, we optimized the electron transport layer, evaluating C₆₀ as well as fullerene derivatives. By this approach and by sputter-depositing an optimized transparent electrode deposited on top of the perovskite cell stack, we made semi-transparent perovskite cells with efficiencies of up to 16 %. By implementing this cell in a monolithic perovskite/Si heterojunction tandem cell and as a result of the improved perovskite layer uniformity, we reached efficiencies of up to 19 % on a cell area larger than 1 cm² and above 21 % for smaller cells.