MOLECULAR DYNAMICS STUDIES OF METALLIC NANOPARTICLES

Abstract

Metal nanoparticles have many desirable electrical, magnetic, optical, chemi-cal, and physical properties. In order to utilize these properties effectively it is neces-sary to be able to accurately predict their size-dependent properties. One common method used to predict these properties is with numerical simulation. The numerical simulation technique used throughout this effort is the molecular dynamics (MD) si-mulation method. Using MD simulations I have investigated various metallic nano-particle systems including gold nanoparticles coated with an organic self-assembled monolayer (SAM), the self-propagating high-temperature synthesis (SHS) reaction of nickel and aluminum nanoparticles, and the mechano-chemical behavior of oxide coated aluminum nanoparticles. The model definition, boundary conditions, and re-sults of these simulations are presented in the following dissertation. In the first material system investigated MD simulations are used to probe the structure and stability of alkanethiolate self-assembled monolayers (SAMs) on gold nanoparticles. Numerous results and observations from this parametric study are pre-sented here. By analyzing the mechanical and chemical properties of gold nanopar-ticles at temperatures below the melting point of gold, with different SAM chain lengths and surface coverage properties, we have determined that the material system is metastable. The model and computational results that provide support for this hy-pothesis are presented. The second material system investigated, namely sintering of aluminum and nickel, is explored in chapter 4. In this chapter MD simulations are used to simulate the kinetic reaction of Ni and Al particles at the nanometer scale. The affect of par-ticle size on reaction time and temperature for separate nanoparticles has been consi-dered as a model system for a powder metallurgy process. Coated nanoparticles in the form of Ni-coated Al nanoparticles and Al-coated Ni nanoparticles are also analyzed as a model for nanoparticles of one material embedded within a matrix of the second. Simulation results show that the sintering time for separate and coated nanoparticles is dependent upon the number of atoms or volume of the sintering nanoparticles and their surface area. We have also found that nanoparticle size and surface energy is an important factor in determining the adiabatic reaction temperature for both systems, coated and separate, at nanoparticle sizes of less than 10nm in diameter. The final material system investigated in chapters 5 and 6 is the oxide coated aluminum nanoparticle. This material system is simulated using the reactive force field (ReaxFF) potential which is capable of considering the charge transfer that occurs during oxidation. The oxidation process of oxide coated aluminum nanoparticles has been observed to occur at a lower temperature and a faster rate than micron sized nanoparticles, suggesting a different oxidation mechanism. From this effort we have discovered that the oxidation process for nanometer sized oxide coated aluminum particles is the result of an enhanced transport due to a built-in electric field induced by the oxide shell. In contrast to the currently assumed pressure driven diffusion process the results presented here demonstrate that the high temperature oxidation process is driven by the electric field present in the oxide layer. This electric field ac-counts for over 90% of the mass flux of aluminum ions through the oxide shell. The computed electric fields show good agreement with published theoretical and experi-mental results. The final chapter includes some important conclusions from this work and highlights some future work in these areas. Future work that is outlined includes ef-forts that are currently underway to analyze the interactions of multiple alkanethiolate coated gold nanoparticles in vacuum and in solvent. Other future efforts are farther out over the horizon and include using advanced computing techniques such as gen-eral purpose graphical processing units (GPGPU) to expand simulation sizes and physical details over what it is currently possible to simulate.