Molecular Dynamics of Disordered Phases
Note: all citations are missing from this HTML except. Download postscript below for references.
Ludwig Boltzmann, who spent much of his life studying statistical mechanics, died in 1906, by his own hand. Paul Ehrenfest, carrying on the work, died similarly in 1933. Now it is our turn to study statistical mechanics.-- David L. Goodstein
It is with well-deserved reservations that we now embark on a
study of disordered phases of silicon with the latest version of EDIP.
Although our experience with cohesive energy curves may foreshadow
problems with the liquid, a metal of greater density than the solid,
the theory behind EDIP certainly addresses metallic bonds and
overcoordination, so we may hope for a reasonable liquid. The success
of the current EDIP for defect structures and the bulk crystal
indicates an exceptionally good description of sp^3 hybrid bonds, so
it would seem likely that the same version would perform well for the
amorphous phase, which is made up of a random network of distorted
tetrahedra. These expectations can only be checked by molecular
dynamics simulations, using a sufficiently large number of atoms to
minimize the influence of periodic boundary conditions and long enough
times to obtain accurate thermal averages. In the first section
we outline some of the techniques required to perform such simulations
on high performance computers.
It is customary in the literature to describe these kinds of studies as computer simulations of a real material (e.g. silicon), which is somewhat misleading. It is more accurate to say that one is exploring the properties of a fictitious material characterized by a particular empirical potential, which may only bear some resemblance to the real material. In this chapter, we examine disordered phases of the current EDIP material through molecular dynamics simulation, including various liquid, glassy and amorphous specimens. The purpose of these of kinds of studies is to explore the transferability of the potential and to develop large-scale simulation techniques. With such knowledge, we may hope to improve the potential and as well as our ability to perform computer experiments to the point where reliable theoretical predictions for real silicon might be possible. The chapter closes with an outlook on the future of EDIP as a transferable model for silicon condensed phases and bulk defects.
Download complete chapter, 42 pages postscript, 455K.
Back to Interatomic Forces in Covalent Solids