Dewetting is an aspect of materials science that directly affects the reliability of microelectronics and the formation of nanostructures with applications in storage, semiconductors, and plasmonics. Dewetting initiates at triple phase lines between film, substrate, and vapor and propagates away from the triple lines through the process of surface self-diffusion due to film curvature towards flat areas of the film (i.e. away from the edge of a film). As edge retraction starts to occur, other instabilities further the process through allowing agglomeration of the retracting film to form islands of high aspect ratio. The solid-state dewetting and long range edge retraction of Au/Ni thin films from a SiO2/Si substrate are driven by the minimization of the free surface and film/substrate interface. While dewetting processes in monolayer thin films have been studied extensively, there are only a few studies involving bilayer thin films and only one that reports long-range edge retraction.

This thesis explores the agglomeration and long-range edge retraction of Au/Ni bilayer films from SiO2/Si substrate. The films were deposited in sequences of Au deposited on top of Ni and Ni deposited on top of Au. As-deposited films were subsequently annealed at 545 ⁰C, 675 ⁰C, and 730 ⁰C. They were characterized by scanning electron microscopy (SEM) and energy dispersive x-ray spectroscopy (EDX) to identify the mechanisms and driving forces for long-range edge retraction. Annealing of Au/Ni bilayer films supported by a SiO2/Si substrate displayed long-range edge retraction occurring, as well as the formation of seven morphologically distinct regions. The formation of each region was reliant on the dominant mechanisms for retraction at play. Interface and volume diffusion were determined to be the dominant mechanisms in the regions closest to the edge of the substrate (regions P6 and P7), while in the transition region P5, interface diffusion was the sole dominating mechanism. Surface diffusion dominated regions P3 and P4. The semi-circular shape in which retraction occurred was indicative of a corner instability, the primary mechanism of long-range edge retraction. This is the first reported instance of a corner instability initiating without a corner, especially for long-range edge retraction. Line-like features of high aspect ratio indicated a secondary retraction mechanism of a fingering instability. The comparison between both deposition sequences revealed the stabilizing effect a native NiOx has on the bilayer film through hindering edge retraction and the critical role that interaction between Au and Ni plays in edge retraction. A uniform area of dark contrast indicating the Si substrate at the P2-P3 interface was observed in all Au/Ni bilayer films. It was proposed that this interface is a result of void formation and nucleation propagated by a thickness variation in region P2 due to the diffusion of Ni to region P3. However, another explanation is the occurrence of a pinch-off event due to the agglomeration of Au and Ni in regions P3 to P5 reaching a critical height. However, additional data is required to ascertain whether that is occurring.

The redox behaviors of nanoscale metal and ceramic materials are affected by various parameters and are important in determining certain properties of the material. This dissertation reports the controlling of oxidation and reduction behaviors in some example metal and ceramic materials at the nanoscale by varying material dimensions, gaseous environments, surface energy densities and application of electric fields and currents. It was suggested in previous studies that iron oxide FeO is thermodynamically unstable under 1000 K with dimensions smaller than 100 nm. In this study, in-situ heating experiments to gradually reduce Fe2O3 nanoparticles under 50 nm and nanochains were conducted in a transmission electron microscope. Electron energy loss spectroscopy and selected area electron diffraction both confirmed previous predictions and also revealed the stabilization of FeO phase in nanochains above a critical length. It provides direct evidence that metal (Ⅱ) oxide with dimensions (particle size) below 100 nm can be stabilized by assembling particles in 1D nanochains. In the case of nickel nanoparticle oxidation, anisotropic growth of nickel oxide nanostructures was observed during in-situ heating of the particles at 800 °C under water vapor atmospheres in an environmental scanning electron microscope. The NiO stoichiometry was confirmed by both energy dispersive X-ray spectroscopy and selected area electron diffraction. Annealing of the nickel particles under different oxygen partial pressures prior to ESEM heating showed that anisotropic NiO growth only takes place at specific locations where local surface energy density is high enough. The results suggested that the oxide growth can be prevented by manipulation of the surface energy by annealing under low oxygen partial pressures. One directional growth of single crystalline nickel oxide nanostructures during nickel particle oxidation is also reported at 650 C and 4 ×10-4 Pa oxygen partial pressure in a transmission electron microscope. In-situ high resolution TEM revealed the layer-by-layer growth of nickel oxide at the Ni/NiO interface while the nickel particle was being consumed. Ledge movement and disconnection migration was observed at the interface that resembled a terrace-ledge-kink growth mechanism. The study demonstrates the applicability of TLK crystal growth mechanism at buried reactive heterophase interfaces. The role of electric current on one directional nickel nanostructure growth in an SEM is also reported in the study. A positive DC bias was applied to a tungsten carbide nanoindenter tip and electric current flow was established by contacting the tip to a nickel particle sitting on a nickel micropillar. Dielectric breakdown of the nickel oxide surface layer was observed prior to nanostructure growth. Nanostructure growth was achieved upon the retraction of the indenter tip. It was demonstrated by both theoretical calculations and finite element modeling that growth was caused by the combination of electromigration and Joule heating. For ceramic materials, the study reports electric field effects on the (100) twist grain boundary core structure of SrTiO3 bicrystals. Nominal field strengths of 50 V/mm and 150 V/mm were applied to the bicrystal in the direction that’s parallel to the grain boundary during high temperature annealing after the formation of the grain boundary. High angle angular dark field imaging displayed different grain boundary structure near the positive and negative electrode side. Electron energy loss spectroscopy and X-ray photoelectron spectroscopy both showed higher oxygen vacancy concentrations, i.e. more local reduction near the negative electrode side. The defect redistribution is caused by the migration of oxygen vacancies to the negative electrode side driven by the external electric field.

Next generation backside imager photodiodes are presently being developed for use in hard x-ray detection and imaging at the National Ignition Facility (NIF) at Lawrence Livermore National Lab (LLNL). NIFs mission is to establish nuclear fusion capabilities, and accurate high-speed diagnostics play a critical role in advancing research. Replacing current silicon-based photodiodes with germanium-based technology allows for utilization of germanium’s improved intrinsic semiconductor properties to create diodes that are thinner and faster. Creating a photodiode with optimized speed and detection limits requires the backside imager geometry in which the substrate is removed, and the bottom of the wafer is used for photon collection. Established silicon etching techniques cannot be applied to germanium due to different fundamental chemical properties thus the selective removal of a germanium substrate is a yet unsolved processing problem.This research project has investigated the use of intentional doping of germanium layers to control and effectively accelerate or arrest electrochemical etching. Research activities were separated into three sections: development of a system and methodology for investigating the electrolytic etch process for germanium, identification of a suitable etchant system, and demonstration of etch selectivity within the determined system. The major result of this PhD project is that under anodic etching conditions with an applied bias of -2V, high-purity intrinsic germanium etches marginally whereas etching of highly boron-doped germanium is accelerated by at least 3 orders of magnitude. A potential processing methodology is proposed that will enable manufacturing of germanium-based photodiodes for future backside imaging devices.