Statistical Dependence of Gate Metal Work Function on Various Electrical Parameters for an n-Channel Si Step-FinFET.

This paper presents a 3-D statistical simulation study of an n-channel Si step-FinFET due to work-function variability depending on grain orientation of metal gate. The statistical fluctuation induced by metal gate granularity on threshold-voltage ( \sigma VT ), ON current ( \sigma I \mathrm{\scriptscriptstyle ON} ), and OFF current ( \sigma I \mathrm{\scriptscriptstyle OFF} ) are estimated for a wide range of channel length and effective fin width for varying average grain size using Technology Computer Aided Design simulator. We investigated the effect of grain on magnitude of variability and also on shapes of various electrical parameters of distribution. The results indicate that \sigma VT decreases for increase in each dimension of the device. The value of \sigma I \mathrm{\scriptscriptstyle ON} , and \sigma I \mathrm{\scriptscriptstyle OFF} reduces as channel length increases. However, \sigma I \mathrm{\scriptscriptstyle ON} and \sigma I \mathrm{\scriptscriptstyle OFF} increase as fin width increases. The distribution of electrical parameters is near to normal for small grain size and becomes bimodal at large grain size. The proposed structure shows excellent behavior in terms of threshold-voltage ( VT ), subthreshold swing, and current ratio ( I \mathrm{\scriptscriptstyle ON}/I \mathrm{\scriptscriptstyle OFF} ) compared with conventional FinFET at high temperature. It is found that the hot carrier effect reduces with increased effective channel width. [ABSTRACT FROM AUTHOR]