Optimising monolithic perovskite tandem photovoltaics for maximum energy yield

Photovoltaic cells are generally optimised for operation at 25 °C under the AM1.5 solar spectrum. This is useful for comparing different technologies, but ultimately energy yield (the energy output across a year of operation) is what we aim to maximise. This is generally estimated using the power conversion efficiency (PCE) of a cell at 25 °C and its temperature coefficient (how PCE changes with temperature). For monolithic tandem cells, this method is less accurate as current matching between each sub cell is required for maximum performance. To understand how to optimise tandems for energy yield, we fabricated perovskite/Si (30.8% PCE) and all-perovskite (25.4% PCE) tandems and tested them over a range of temperatures. Opposing temperature dependence with bandgap - perovskite bandgaps tend to widen with increasing temperature, whereas Si narrows - results in a worse temperature coefficient for perovskite/Si (-0.23%/°C) compared to all-perovskite tandems (-0.17%/°C). Increasing the Cs concentration at the perovskite A-site reduces the bandgap widening effect, resulting in an improved temperature coefficient (-0.19%/°C). Surprisingly, when estimating energy yields using real-world data, this improved temperature coefficient does not translate to increased yields. Instead, we find that the opposing temperature/bandgap correlation of perovskite/Si tandems is an advantage as the solar spectrum is blue-shifted compared to AM1.5 when cells are warmest and operate at their peak power output. Thus, we expect an increase in energy yield of up to 3% for optimised perovskite/Si tandems compared to other tandem technologies with the same PCE. Significantly, this equates to a full PCE point increase under standard conditions for optimised (> 30%) perovskite/Si tandems.