Design trade-offs in InGaN top cells for high-efficiency four-terminal InGaN/Si solar cells

This study presents a detailed modeling and optimization analysis of mechanically stacked four-terminal (4 T) indium gallium nitride/silicon (InGaN/Si) tandem solar cells using SCAPS-1D coupled with the Lion Optimization Algorithm (LOA). Four InGaN top-cell architectures (PN, PIN, PNN and PPN) were independently optimized for thickness, doping and In composition to maximize the photovoltaic (PV) performance metrics (V oc , J sc , FF and PCE). Single-objective optimization yielded maximum single-junction PCEs of approximately 16.6% for the PPN InGaN configuration and 21.6% for the Si bottom cell. A weighted multi-objective optimization approach was then applied to generate practical compromise designs suitable for 4 T stacking. Empirically based absorption coefficients and Beer–Lambert transmission calculations were used to determine the transmission spectra for the bottom Si cell and to estimate combined tandem PCEs. Optimized mechanically stacked designs, particularly PPN/Si, achieved combined PCEs of up to 33.3% under AM1.5G illumination. The findings demonstrate that device architecture (intrinsic and field layers) and In fraction strongly influence spectral splitting and photon collection. This work further shows that incorporating realistic structural constraints with LOA enables the design of high-efficiency 4 T tandem solar cells.