Novel Monolithic Three-Terminal Tandem Solar Cells Based on Antimony Chalcogenide Absorbers

Among emerging photovoltaic absorbers, quasi-1D (Q-1D) materials such as Sb₂(S,Se)₃ are viewed as particularly promising owing to their high degree of electrical and chemical tunability. Benefiting from bandgap controllability within the 1.2eV-1.8eV range controlled by the S to Se ratio, the unidimensional ribbon-like structure theoretically allows to tailor the film conductivity in an anisotropic manner while the Van der Waals regions between ribbons should permit a higher degree of control in doping and elemental intercalation. Additionally, the comparatively low processing temperature of this class of solar cells and the feasibility of both substrate and superstrate devices makes the Sb₂(S,Se)₃ system a prime candidate for being the first monolithic, full chalcogenide tandem solar cell technology.<br/>In this work, we evaluate the feasibility of such device. Starting with a combination of quantitative optical and electrical modelling approaches with a realistic set of parameters from in-lab characterizations, we investigate various possible tandem configurations using an Sb₂S₃ top subcell and an Sb₂Se₃ bottom subcell. We particularly emphasize original three-terminal designs in which the top and bottom cell share a middle Transparent Conducting Oxide (TCO) contact. Such architecture allows both subcells to operate individually (no current matching necessary) while remaining compatible with a monolithic fabrication process. Taking advantage of the process flexibility inherent to antimony chalcogenide technologies, different experimentally possible designs are studied, with either both diodes in the same direction (n-p/TCO/n-p) or in opposite directions (p-n/TCO/n-p), each material stack with its own merits. The optical modelling, performed using the transfer matrix method, allows to identify the most critical material parameters to optimize for maximizing the transparency of the top subcell below its bandgap, with a particular emphasis on TCO and ohmic contact transparency. In that context, using an MoO₃ interlayer is found vastly preferable compared to using MoSe₂ for contact ohmicity between the top absorber and the shared TCO. Similarly, using high mobility TCOs is found to be critical in improving infrared transparency. By combining quantitative optical modelling with an accurate electrical modelling of both subcells, we obtain a reliable evaluation of monolithic three-terminal tandem solar cells based on Sb₂S₃ and Sb₂Se₃ absorbers. While no current matching is necessary, we show that when considering an architecture in which the p-n junction of a subcell is located at its back interface, it is critical to nevertheless optimize the absorber thickness to maximize carrier collection. In such case, it is demonstrated using state-of-the-art material parameters and proper architecture optimizations that various configurations lead to a tandem efficiency of 11%. It is additionally shown that if quenching interface recombination in both subcells (originating from defect “D3” reported in the literature) using a front surface field, the three-terminal tandem efficiency can realistically overcome the 20% threshold (up to 23%), a value well within range of other more mature technologies. To conclude this study, an experimental proof of concept three-terminal tandem device is presented using a p-n/TCO/n-p architecture, where the p-n bottom cell is fabricated in substrate configuration by combining metallic precursor sputtering and reactive annealing, and the n-p bottom cell is fabricated by Atomic Layer Deposition. Such device is to the best of our knowledge the first full chalcogenide monolithic tandem solar cell, and represents the first step toward thin film multijunction becoming a relevant field of interest for future low-cost, high efficiency and versatile photovoltaic devices.