Determination of the growth mechanisms of gamma-TiH in alpha-Ti

High resolution transmission electron microscopy (HRTEM), image simulation and image analysis were used to examine the atomic structure of the (01 10) {dollar}alpha{dollar}-Ti/{dollar}gamma{dollar}-TiH interface and ledges at the interface. Analysis of the experimental images reveals that the experimentally observed interface can be simulated and that ledges at the interface contains a displacement of {dollar}{lcub}1over 3{rcub}langle 2overline 1 10rangle{dollar}, consistent with the presence of two {dollar}{lcub}1over 3{rcub}langle 10bar 10rangle{dollar} partial dislocations, each of which displays an {dollar}{lcub}1over 6{rcub}langle 2overline 1 10rangle{dollar} edge component in projection in the HRTEM images. An edge-on ledge model was created based on the transformation dislocations associated with the ledge in the experimental images, and an inclined ledge model can be created by stacking individual edge-on ledge models with different ledge length. The good agreement between the simulated images of the inclined ledge model and the experimental images of the hydride ledges shows that hydride ledges in general are inclined.; The difference between the matrix and hydride plasmon peaks in EELS allows plasmon imaging of hydrides, and sub-nanometer spatial resolution was obtained in plasmon imaging of the {dollar}alpha{dollar}-Ti/{dollar}gamma{dollar}-TiH interface. It was determined that the hydride was fully formed with the chemical composition of TiH up to the interface and the ledge fronts. This indicates that diffusion of hydrogen is required in the transformation process. The results from the in situ plasmon imaging experiments there was no change in the crystal structure in front of the ledge. Therefore the {dollar}alpha{dollar}-Ti {dollar}to gamma{dollar}-TiH transformation process requires a sufficient concentration of hydrogen at the ledge front.; In situ HRTEM experiments using the electron beam as a heating source were also conducted. Measured ledge velocities were compared with a simple model of ledge migration according to diffusion controlled kinetics. The experimental results indicate that migration of the {dollar}gamma{dollar}-TiH ledge may be interfacially controlled, but not martensitic. Three possible rate-limiting factors were discussed: (1) the large volume strain ({dollar}sim{dollar}15%) associated with the transformation, (2) hindrance to ledge movement as a result of the intersection between hydride plate and foil surface, and (3) a low supersaturation of hydrogen in the samples. All three factors require some type of nucleation event in order for ledge migration to occur.