Analyze of process parameter in 19 nm WSi2/TiO2 N-channel MOSFET using statistical modelling

Amira Emyliana, Zailan (2017) Analyze of process parameter in 19 nm WSi2/TiO2 N-channel MOSFET using statistical modelling. Project Report. UTeM, Melaka,Malaysia. (Submitted)

Text (24 Pages) Analyze of process parameter in 19 nm WSi2TiO2 N-channel MOSFET using statistical modelling.pdf - Submitted Version Download (2MB)

Abstract

This project investigates the impact of process parameter on threshold voltage (VTH),drive current (IoN) and leakage current (loFF) for 19nm WSi2/Ti02 NMOS device.using statistical modelling.The statistical methods that were used are 2k-factorial design and Taguchi method.The four process parameter,namely oxide growth temperature,VTH implant energy,pocket halo implant dose and compensate halo implant will be investigated and adjusted to improve the results.The simulated of the device was performed by using ATHENA module.While the electrical characterization of the device was implemented by using ATLAS module.These two modules will be combined with 2k-factorial and Taguchi Method to aid design and optimize the process parameters. The most effective process parameter with respect to threshold voltage,drive current and leakage current were chosen depend on the percentage of the factor effect on S/N ratio that indicate the relative power of factor to reduce variation.In this research,optimum condition for electrical characteristics have been compared using two methods to choose the best method for optimization process.The most dominant or significant factors for S/N Ratio are pocket halo implant dose and compensate implant dose.Meanwhile,the S/N Ratio values of VTH,IoN and loFF after the optimization approaches for array L9 is 4 1.30dB,55.70dB and -14.17dB respectively.In L9 experiments,VTH,IoN and loFF values for NMOS device after optimizations approaches are 0.530V, 622.667uA and 2.776pA respectively.The results obtained are closer to ITRS 2013 prediction.As conclusions,Taguchi Method was observed to be the most suitable method to be implemented in statistical modeling of 19nm WSi2/Ti02 NMOS device.

Actions (login required)

View Item