Abstract:
Silicon, the second most abundant element in the earth’s crust, is ubiquitous
in the form of silica and silicates in the natural world. In the elemental form,
it is an essential component of the semiconductor technology. It was first
prepared in its amorphous form by J.J. Berzelius and later the crystalline
form by H.E. Sainte-Claire Deville [185] in the 1800’s. The crystalline and
amorphous solid are the two most familiar forms of silicon, which have been
studied extensively. The crystalline form of silicon is a tetravalent semiconductor (as is the amorphous solid) and upon melting at 1687K at ambient
pressure, transforms to a metallic liquid with higher coordination number,
around 6. Liquid silicon is relatively less studied, given the elevated temperatures at which it exists. Nevertheless, it has been a subject of substantial
experimental, theoretical and computational investigation, both at temperatures above the melting temperature, and in the supercooled and stretched
(negative pressure) states. The investigations of the metastable liquid (see
FIG. 1.1) have been motivated, as this thesis seeks to demonstrate, by fundamental questions regarding (i) the eventual fate of metastable liquids upon
deep undercooling and stretching, (ii) the interest in the possibility of a novel
transition between two distinct liquid forms in a class of “tetrahedral” liquids to which silicon belongs, and (iii) the role of the thermodynamics of
metastable liquid states on the kinetics of phase transformations, particularly to the crystalline state.
