Development of TOPCon-based Solar Cell Architectures : From Single-Junction Silicon Cells to Perovskite-Silicon Tandems

The global energy transition requires photovoltaic technologies capable of surpassing the thermodynamic efficiency limit of single-junction silicon cells (29.4%). While crystalline-silicon tunnel-oxide passivating-contact (TOPCon) solar cells currently dominate the market, their efficiencies are saturating. To overcome this barrier, the industry is pivoting toward perovskite-silicon tandem architectures, which have recently demonstrated efficiencies approaching 35%. This thesis aims to bridge the gap between industrial c-Si manufacturing and next-generation tandems by addressing critical challenges: simplifying TOPCon-based cell fabrication, adapting the architecture for monolithic perovskite-silicon tandem integration, and developing sustainable interconnection layers. First, we investigated the simplification of n-type TOPCon manufacturing. We developed a streamlined co-annealing process using a PECVD boron-doped silicon oxide layer as a solid-state diffusion source. This approach enables simultaneous front emitter formation and rear poly-Si crystallization in a single thermal step at 900°C. Tuning the trimethylboron flow controlled doping but revealed a fundamental limitation: low flows minimized Auger recombination but resulted in shallow junctions (<100 nm) shunted by metallization (metal recombination current >8000 fA/cm²), while higher flows introduced severe intrinsic recombination. While a proof-of-concept efficiency of 21.0% was achieved, this trade-off necessitates alternative architectures for high-efficiency applications. Second, we investigated double-sided passivating contact solar cells (TOPCon²) with firing-through metallization. We demonstrated that thick front poly-Si layers require localization to avoid parasitic absorption. While localized contacts enabled efficiencies exceeding 21.0%, simulations indicate limited gains over state-of-the-art LECO-TOPCon, suggesting the complexity of patterning outweighs benefits for single-junction applications. Consequently, we focused on developing a TOPCon² bottom cell tailored for monolithic tandems. We identified that standard industrial hydrogenation damages nanotextured surfaces during etching. We engineered a low-temperature hydrogenation process using a thin (10 nm) AlOx layer, allowing rapid removal without contact degradation. Additionally, we characterized damage induced by sputtering the ITO recombination junction. We identified 300°C as the optimal curing temperature to recover passivation without increasing contact resistivity. Integrating these processes enabled monolithic perovskite/TOPCon tandems with champion efficiencies of 30.2% on flat wafers and 31.3% on front-nanotextured bottom cells. Finally, we addressed optical limitations and sustainability by eliminating the indium-based TCO recombination junction. We replaced ITO with a silicon-based tunnel junction using boron-doped nanocrystalline silicon (nc-Si:H(p)), developed via low-temperature PECVD. Optimizing CO2 plasma pre-treatment and hydrogen dilution yielded a conductive, thermally stable layer. Upon integration, we found standard SAMs (e.g., Me-4PACz) failed on nc-Si due to the lack of a substrate-dependent work-function shift. We engineered a "mixed-SAM" formulation to restore the necessary dipole. Integrating this TCO-free junction yielded efficiencies of 31.2% on SHJ and 30.7% on TOPCon² bottom cells. EQE analysis confirmed a current gain of +0.30 mA/cm² due to eliminated parasitic absorption.