Page 195 - Semiconductor For Micro- and Nanotechnology An Introduction For Engineers
P. 195

Transport Theory
                             duce the concept of scattering. This term will obtain more meaning in
                             Chapter 7, where we look at the interactions between the various sub-
                             systems.
                Chapter Goal The goal of this chapter is to show how to describe transport phenomena
                             and how to compute the transport of electrons through a semiconductor
                             device.

                Chapter      We start with the Boltzmann transport equation, describing its terms, and
                Roadmap      show how it should be modified for particular particle types. Zooming in,
                             we first describe local equilibrium. Next, we see how local equilibrium
                             and the global balance equations of thermodynamics are related. This
                             leads us to a number of models for classical and semi-classical transport
                             of charge carriers in silicon, including the very successful drift-diffusion
                             equations. Finally, we look at numerical methods used to solve the trans-
                             port equations.





                             6.1 The Semi-Classical Boltzmann Transport
                                 Equation
                             The semi-classical Boltzmann transport equation (BTE) gives a descrip-
                             tion of the electronic system in terms of single particle distribution func-
                                     ,,
                                   (
                             tions  f px t)  , i.e., a time-dependent density in a six-dimensional
                             single particle phase space. A single particle is identified by its three spa-
                             tial coordinates and three momentum values. For the momentum of a free
                             particle we know that  p =  —k =  mv  . In this case we could write the
                                                                 (
                                                                   ,,
                             distribution function in terms of velocities  f vx t)  . From the discus-
                             sion in Section 3.3.3 we know that this simple relation does not hold in a
                             semiconductor in general.  Though we shall restrict our discussion to
                             energies near the band extrema, where  p =  m∗ v   holds, with the effec-
                             tive mass  m∗  , we prefer to take the wave vector   as the independent
                                                                      k
                             variable. This will remind us that in the case of electron we are dealing
                             with wave phenomena as well as with particle properties. A more formal


                192          Semiconductors for Micro and Nanosystem Technology
   190   191   192   193   194   195   196   197   198   199   200