Optimal Doping Profile of MOSFETs Using Geometric Programming

Abstract

This study employs a novel geometric programming technique to optimize the doping profile of metal semiconductor oxide field effect transistor (MOSFET) subject to specified device characteristics, such as the threshold voltage, the on- and off-state currents and the subthreshold swing. The optimal design of doping profiles for MOSFET devices is for the first time formulated as a geometric programming problem. This special type of optimization problems can be transformed into a convex optimization problem, and therefore can be solved globally and efficiently. For targeting a designed threshold voltage as well as obtaining a high transconductance, the doping distribution of an examined MOSFET device is extracted. Consequently, various constrains of DC characteristics are estimated to obtain the desired doping profile. The approach provides an alternative way to accelerate the design of doping profile process and benefits the manufacturing of MOS devices.