Photovoltaic (PV) module production in the multi-TW scale is required for cost-efficient climate change mitigation. In the coming decades, a substantial fraction of this market may be provided by perovskite solar cells (PSC) and perovskite-based tandem concepts. Although the thin-film perovskite PV concepts require minute layer thicknesses, the huge required annual PV production rates of thousands square kilometers imply material demands in the range of kilotons. This raises the questions which materials and processes chemicals are sufficiently abundant, what is the impact of the CO2e emissions, and how the accumulating waste streams can be recycled.

First, we discuss the CO2e emissions of state-of-the art silicon as well as perovskite PV technologies. We show that, although the carbon footprint of silicon PV technologies is much lower than that of fossil fuel-based energy technologies, even with continued technological learning, the CO2e emissions of the PV industry may require 4 to 11% of the remaining carbon budget to limit global warming to 1.5°C. With perovskite PV technologies, the CO2e emissions can be further reduced to the lower boundary represented by the glass substrate.

Moreover, we present a comprehensive quantitative assessment of the material demand for multi-TW scale perovskite PV production. The supply criticality is assessed by comparing inorganic and synthetic material demands with current production rates and by estimating the scalability of material production. We find that scaling perovskite PV production is feasible from a material perspective and that the it is valid to claim that PSC are “made from abundant materials”. However, a range of materials which are commonly applied in highest-efficient PSC have been identified as critical in supply: Indium and gold used in transparent and opaque electrodes and interconnection layers must be replaced. Current cesium mining is several orders of magnitudes below the required production. This finding is especially critical for the research on stable high-bandgap inorganic perovskites. Furthermore, with the exception of PEDOT:PSS, the synthesis routes of the organic hole transport materials are currently incompatible with industrial upscaling. In contrast, the industrial production of most synthetic nanoparticulate materials has sufficient maturity. The supply of organic solvents is also not critical.

Finally, the foreseeable massive material waste streams mandate that researchers adapt a design-for-recycling thinking already in early stages of technology development. Possibilities to realize this are demonstrated by an approach to remanufacture fully encapsulated perovskite solar mini-modules.

Overall, we show that besides the improvement of efficiency and stability, the perovskite PV research community should also focus on the various aspects of long-term sustainability considerations.