|M.Sc Student||Chachamovitz Doron|
|Subject||The Activation Parameters for Heterogeneous|
Dislocation Nucleation in Mo Nanoparticles
|Department||Department of Mechanical Engineering||Supervisor||Professor Dan Mordehai|
|Full Thesis text|
There is an increasing use of objects of nanoscale dimension, such as nanoparticles, as basic building blocks of nano-mechanical systems. The strength of these objects, which are initially pristine from dislocations, imperfections in the lattice structure, is much higher than their bulk counterparts, due to the need to nucleate dislocations on their surfaces. The process of nucleating dislocations, which is a thermally activated process, depends on local stresses. Estimating the energy barriers and rates of surface nucleation is still a challenging task, both computationally and experimentally. In this thesis, we study the strength of molybdenum (Mo) nanoparticles under compression, and propose a computational scheme to examine both how temperature and size affect the nucleation parameters.
We combined here between classical nucleation theory and molecular dynamics (MD) simulations to calculate the activation energies and volumes to nucleate dislocations from vertices of faceted nanoparticles. Mo nanoparticles, which have a body-centered cubic (BCC) lattice structure, were chosen as a model system, also owing to the reliable interatomic potentials existing for this material. The stress to nucleate dislocations was calculated for nanoparticles of various sizes and at different temperatures. The simulation was repeated several times for each size and temperature and the nucleation stresses were found to be distributed around an average value. This average value, was found to decrease both with increasing temperature and size.
With the help of classical nucleation theory, a model is proposed, to extract the activation parameters for dislocation nucleation during compression from the average nucleation stress and the distribution around it. The model was applied on the simulation results, and it was found that the activation volume is of the orders of a few unit cells. Moreover, the activation volume was found to increase with the size of the nanoparticles. Fitting the distribution function from the model to the MD simulation results uncovered the stress dependence of the activation energy and volume, as well as the nucleation rates. The latter was employed to calculate the activation entropy of the system. Finally, we found from the activation parameters the melting point at the vertices of the nanoparticles. The results are in good agreement with previous results obtained with free-energy techniques, which are more complex and tedious than our proposed method.