Preface
This thesis is an account of the work carried out by the author during the period October 1994 to June 1997,
under the supervision of Dr. G.A.D.Briggs and Professor D.G.Pettifor. It has not been previously submitted for a
degree at this or any other university. The work of others, which has been drawn upon, is acknowledged as such
where used. Parts of this thesis have been published or are in press, and a list of such parts is given below.
Owen J.H.G., Bowler D.R., Goringe C.M., Miki K. and Briggs G.A.D. (1995) Surf. Sci. Lett. 341, L1042.
Bowler D.R. and Goringe C.M. (1996) Surf. Sci. Lett. 360, 489.
Owen J.H.G., Bowler D.R., Goringe C.M., Miki K. and Briggs G.A.D. (1996) Phys. Rev. B 54, 14153.
Owen J.H.G., Bowler D.R., Goringe C.M., Miki K. and Briggs G.A.D. (1997) Surf. Sci. Lett. 382, L678.
Bowler D.R., Aoki M., Goringe C.M., Horsfield A.P. and Pettifor D.G. (1997) Modelling Simul. Mat. Sci. Eng. 5, 199.
Owen J.H.G., Miki K., Bowler D.R., Goringe C.M., Goldfarb I. and Briggs G.A.D. (1997a) Surf. Sci. In press.
Owen J.H.G., Miki K., Bowler D.R., Goringe C.M., Goldfarb I. and Briggs G.A.D. (1997b) Surf. Sci. In press.
Goringe C.M., Bowler D.R. and Hernández E.H. (1997) Rep. Prog. Phys. In press.
Goringe C.M. and Bowler D.R. (1997) Phys. Rev. B, In press.
A theoretical study of gas source growth of the Si(001) surface
D.R.Bowler, Wolfson College
Thesis submitted for the degree of Doctor of Philosophy
Trinity 1997
Abstract
The growth of the Si(001) surface from gas sources such as disilane is technologically important, as well as
scientifically interesting. The aspects of growth covered are: the clean surface, its defects and steps; the action of
bismuth, a surfactant; the diffusion behaviour of hydrogen in different environments; and the entire pathway for
formation of a new layer of silicon from adsorption of fragments of disilane to nucleation of dimer strings. The
theoretical methods used, density functional theory and tight binding, are described. Four linear scaling tight
binding methods are compared. The construction of the tight binding parameterisations used is also
explained.
The structure of the most common defect on the Si(001) surface is identified by comparison of the electronic
structure with scanning tunneling microscopy (STM) images. The energy and structure of steps is
calculated, and their kinking behaviour is modelled, achieving good agreement with experimental
results.
Two unusual features which form when bismuth is placed on the surface and annealed are investigated. The first
has possible applications as a quantum wire, and its structure and growth are described. The second
relates to a controversial area in the field; a structure is proposed which fits all available experimental
evidence.
The behaviour of hydrogen is vital to understanding growth, as large amounts are deposited during disilane
growth. After validating the tight binding parameterisation against DFT and experiment for the system of a single
hydrogen diffusing on the clean Si(001) surface, the barriers for diffusion on the saturated surface, down a step
and away from a defect are found, and prove to be in good agreement with available experimental
data.
The pathway for the formation of a new layer of silicon from disilane is described step by step, giving barriers
and structures for all events. The interaction with experiment is highlighted, and demonstrates that great benefit
accrues from such close work, and that the atomistic modelling techniques used in the thesis produce results in close
agreement with reality.