Drift-Diffusion Monte Carlo Simulation for Predictive Photon Detection Probability in FSI SPADs across UV–VIS–IR

Abstract — An enhanced drift-diffusion Monte Carlo framework is presented for accurately modeling the photon detection probability (PDP) of front-side illuminated (FSI) single-photon avalanche diodes (SPADs) across the UltraViolet-Visible-Infra-Red (UV–VIS–IR) spectrum. This approach combines wavelength-dependent absorption modeling with measured quantum efficiency (QE) to estimate optical losses, and, unlike existing electric-fieldbased models, computes avalanche triggering probability based on carrier trajectories, capturing the contributions of deep-penetrating photons at longer wavelengths. Experimental validation across a range of temperatures, voltages, device pitches, and fabrication processes demonstrates that PDP prediction errors remain below 5% across the entire UV–VIS–NIR spectrum accurately capturing observed QE–PDP discrepancies. Additionally, PDP values exceeding 67% in the UV–VIS range were achieved for small-pitch devices. This framework provides a robust and scalable simulation tool for predictive SPAD design optimization without parameter fitting.