A Novel Gaussian Doping Profile Model for High-Efficiency TCS Prediction in Piezoresistive Sensors

The temperature coefficient of sensitivity (TCS) is a key performance metric for piezoresistive sensors, but accurately predicting TCS in sensors fabricated using multiple ion implantation steps remains challenging due to simulation and measurement limitations. This paper presents a Gaussian doping profile model for high-efficiency TCS prediction, reducing TCAD simulation error by 32.32%. When applied to a measured doping profile within only 1/6 of the junction depth, the error decreases to 7.4%. This reduces reliance on characterization techniques like SIMS, lowering costs and improving modeling accuracy. Based on the model, this study first identifies that the dopant concentration variation rate is the dominant factor influencing TCS under the non-uniform doping profile, offering valuable insights for design optimization.