Design Window for SHJ Cells: Joint Impact of Base Thickness and Doping Under AM1.5G

We numerically map the joint impact of base thickness (<i>e</i>) and donor density (<I>N<SUB>D</SUB></I>) on silicon heterojunction (SHJ) cells under AM1.5G using SILVACO ATLAS (drift–diffusion with SRH/Auger and field/concentration dependent mobilities). Optics is treated by 2D specular ray tracing (no texturing), so results constitute a conservative baseline at small e. To isolate <i>e</i> and <I>N<SUB>D</SUB></I>, the a Si:H/c Si interface is held fixed across parameter sweeps. We identify an absorption–collection trade off: <i>J<sub>sc</sub></i> increases from thin to moderate <i>e </i>and then saturates or declines; <i>V<sub>oc</sub></i> decreases with increasing <i>e</i> and high <I>N<SUB>D</SUB></I> due to enhanced recombination. The fill factor peaks at small <i>e</i> under low to moderate <I>N<SUB>D</SUB></I>. Efficiency exhibits a robust optimum at moderate thickness (e.g., <i>e</i> ≈ 120-160<i>μm</i>) and intermediate <I>N<SUB>D</SUB></I>, whereas heavy doping shifts the optimum but ultimately degrades <i>V<sub>oc</sub>/FF</i> via Auger (and, in extended models, band gap narrowing). From a design standpoint, we delineate a practical window that balances resistivity and recombination while avoiding heavy doping. Limitations: absence of light trapping and fixed interface, mean our absolute metrics are conservative, but trends and optima are robust. Planned extensions include Lambertian/textured optics, interface sweeps, and calibrated BGN/contact models to raise absolute values without altering the identified trade offs.