As the energetic demands of our society keep rising, any emerging technology faces the challenge of matching the commercial benefits presented by fossil fuels in terms of energy storage and mobility. Since a vast part of the modern energetic infrastructure already relies on fuels, their production in a sustainable manner stands out as an obvious solution. Therefore, the light-driven conversion of small molecules such as water and CO₂ into so-called solar fuels (e.g. H₂, CO) represents an attractive alternative for simultaneous energy harvesting and storage.^[1,2] While great progress has been made in the development of light absorbers maximizing the solar spectrum coverage, their integration with catalysts into photoelectrochemical (PEC) devices for fuel production still poses challenges. Beside the performance and stability of common PEC prototypes, their scalability and choice of catalyst are also major factors which must be considered for commercial applications.

In this work, we address those issues by developing approaches to synthesize and characterize up-scaled PEC devices, which are able to produce fuels autonomously in the absence of external bias. To overcome the overpotential losses of electroreduction, we introduce state-of-the-art photocathodes obtained by protecting triple cation mixed halide perovskite solar cells with a Field’s metal encapsulant. Their PEC performance is benchmarked using a platinum nanoparticle catalyst for proton reduction, reaching photocurrents as high as -14 mA cm^-2 at 0 V against the reversible hydrogen electrode (RHE) under 1 Sun irradiation.^[3,4,5] By combining the perovskite photocathodes with robust BiVO₄ photoanodes, tandem PEC devices can be obtained which sustain unassisted water splitting at a solar-to-hydrogen conversion of approximately 0.3%. The PEC devices present remarkable stabilities of up to 20 h under operation in an aqueous medium,^[4,5] revealing key general insights for the encapsulation of perovskite optoelectronic devices.

To investigate the scalability of our “artificial leaves”, we prepare devices of sizes up to 10 cm² which reveal a similar performance to their 0.25 cm² counterparts. The PEC tandems are characterized in a versatile 3D-printed PEC cell, which can accommodate a wide array of samples due to its modular design.^[5] The potential for further device up-scaling is revealed by fabricating 25 and 300 cm² doped BiVO₄ panels from bismuth (transition metal) polyoxovanadate single-source precursors.^[6] The overall findings are applicable to a wide range of photoelectrochemical systems employing various photoabsorbers.^[7] Looking beyond water splitting, we will also discuss our recent progress on the development of PEC devices that can couple the more challenging CO₂ reduction to water oxidation, with the ultimate goal of contributing towards a circular carbon economy via photoelectrocatalysis.