An improved empirical model for a semiconductor’s velocity-field characteristic applied to gallium arsenide

Abstract

An effective empirical model for the velocity-field characteristic associated with a semiconductor offers the community of semiconductor device engineers with a tool that will ultimately allow them to relate key features of a semiconductor’s velocity-field characteristic with the resultant device performance. Noting the limitations of previously devised empirical models for a semiconductor’s velocity-field characteristic, we have developed an alternate approach. In particular, building upon the empirical velocity-field characteristic model of Eastman et al. related to high-field domains (Eastman et al., 1980), we develop an approach wherein the model’s internal adjustable parameters may be directly tied with the velocity-field characteristic’s observable features. We then apply this approach to the specific case of gallium arsenide, a contrast with results gleaned from Monte Carlo electron transport simulations being used to show that the resultant fits are indeed quite satisfactory. A comparison with other empirical velocity-field characteristic modeling approaches is offered, our approach being shown to be more effective than the other ones in several regards. Finally, within the framework of this empirical model, a means of evaluating the maximum value of the negative differential mobility associated with a given velocity-field characteristic is provided, such a value being of interest to those who wish to draw upon the aforementioned high-field domain effects.