Loughborough University
Browse
14981.pdf (20.14 MB)

Atomistic interactions of clusters on surfaces using molecular dynamics and hyper molecular dynamics

Download (20.14 MB)
thesis
posted on 2010-10-22, 14:56 authored by Carlos F. Sanz-Navarro
The work presented in this thesis describes the results of Molecular Dynamics (MD) simulations applied to the interaction of silver clusters with graphite surfaces and some numerical and theoretical methods concerning the extension of MD simulations to longer time scales (hyper-MD). The first part of this thesis studies the implantation of clusters at normal incidence onto a graphite surface in order to determine the scaling of the penetration depth (PD) against the impact energy. A comparison with experimental results is made with good agreement. The main physical observations of the impact process are described and analysed. It is shown that there is a threshold impact velocity above which the linear dependence on PD on impact energy changes to a linear dependence on velocity. Implantation of silver clusters at oblique incidence is also considered. The second part of this work analyses the validity and feasibility of the three minimisation methods for the hyper-MD simulation method whereby time scales of an MD simulation can be extended. A correct mathematical basis for the iterative method is derived. It is found that one of the iterative methods, upon which hyper- NID is based, is very likely to fail in high-dimensional situations because it requires a too expensive convergence. Two new approximations to the hyper-MD approach are proposed, which reduce the computational effort considerably. Both approaches, although not exact, can help to search for some of the most likely transitions in the system. Some examples are given to illustrate this.

History

School

  • Science

Department

  • Mathematical Sciences

Publisher

© C.F. Sanz-Navarro

Publication date

2002

Notes

A Doctoral Thesis. Submitted in partial fulfillment of the requirements for the award of Doctor of Philosophy of Loughborough University.

EThOS Persistent ID

uk.bl.ethos.247919

Language

  • en

Usage metrics

    Mathematical Sciences Theses

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC