A New CMOS Simulation Template: Addressing Advanced Diffusion and Annealing Effects
1. Introduction
As the scaling of CMOS devices continues, the accuracy requirements for simulating ion implantation and diffusion processes have become increasingly stringent. Traditional diffusion models, such as “Fermi” model, assumes no excess of point defects and are no longer sufficient due to the increasing complexity of defect interactions and their impact on dopant behavior. To achieve the desired performance in modern nanoscale devices, it is essential to consider the impact of defects on diffusion behavior and to apply corrections for phenomena like Transient Enhanced Diffusion (TED) and Oxygen Enhanced Diffusion (OED) to ensure accurate simulation results. Moreover, as the scaling progresses, junction depths become shallower and doping zones are smaller. To achieve this, advanced techniques like cocktail co-implantation, along with rapid annealing methods, such as Spike and Laser annealing, are being introduced [1][2][3]. However, traditional models often fail to accurately capture these new effects, leading to discrepancies between simulated and real-silicon processes.
To address these challenges, we have revisited and refined all relevant models in Victory Process for CMOS process simulation, introducing a new CMOS simulation framework. This model not only accurately captures the effects of ion implantation, diffusion, and defects, but also offers flexible model options to balance simulation speed and accuracy. In this paper, we will introduce our new CMOS simulation model and demonstrate its practical use and calibration through several case studies.
