研究目的
To develop an accurate analytical model for the standard deviation of threshold voltage (σVth) induced by random dopant fluctuation (RDF) in fully depleted silicon-on-insulator (FD-SOI) MOSFETs, considering both dopant number and position fluctuations, and to validate it against numerical simulations.
研究成果
The proposed analytical model for σVth in FD-SOI MOSFETs, incorporating both dopant number and position fluctuations, shows good agreement with Sentaurus TCAD simulations across various device parameters. Key findings include σVth being proportional to front gate oxide thickness and the square root of the product of channel doping concentration and SOI film thickness, and inversely proportional to channel surface area. The model aids in predicting RDF effects and suggests design strategies like reducing silicon film thickness and doping concentration to minimize variations.
研究不足
The model relies on empirical parameters like Rp, which may vary with process technology and require calibration through simulation. The use of 2D simulations instead of 3D might not capture all effects, though it is justified for computational efficiency. The study is limited to FD-SOI MOSFETs and may not generalize to other device types.
1：Experimental Design and Method Selection:
The study uses an analytical modeling approach based on a short-channel threshold voltage model derived using a quasi-two-dimensional approach and Taylor function simplification. Numerical simulations are performed using the Monte Carlo method with Sentaurus TCAD to validate the model.
2：Sample Selection and Data Sources:
Simulated FD-SOI MOSFET devices with varying parameters such as channel length (Lg), channel doping concentration (NA), SOI film thickness (tsi), front gate oxide thickness (tox), and buried-oxide thickness (tbox). Data is generated using MATLAB for Poisson distributions and random dopant placements.
3：List of Experimental Equipment and Materials:
Sentaurus TCAD software for device simulation, MATLAB for data generation and analysis. No specific physical equipment is mentioned; the work is computational.
4：Experimental Procedures and Operational Workflow:
The channel is divided into grids; dopant numbers follow Poisson distribution, and positions are randomized. Three simulation cases are considered: total variations (number and position), number fluctuations only, and position fluctuations only. Threshold voltage is extracted using linear extrapolation of IDS-VGS curves at VDS = 50 mV.
5：Data Analysis Methods:
Statistical analysis of simulation results (300 samples per case) to compute standard deviations. Comparison between analytical model predictions and simulation data using fitting functions and error analysis.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
