[pt] Neste trabalho foi estudada a deformação mecânica em filmes de GaN por nanoindentação. Um nanoindentador foi usado para induzir a nucleação de defeitos mecânicos na superfície das amostras de forma controlada. A morfologia das indentações e a microestrutura dos defeitos foram estudados com o uso da microscopia de força atômica e microscopia eletrônica de transmissão . Os resultados mostraram que nos estágios iniciais de deformação, o processo de nanoindentação promove o escorregamento em escala atômica de planos cristalinos que pode ser revertido se a carga é removida. Se a carga for aumentada ainda mais, a partir de uma tensão crítica, ocorre um grande evento pop-in com o escorregamento dos planos 1101, 1122 e 0001 produzindo então deformação plástica irreversível. A influência dos dopantes na deformação mecânica foi estudada e os resultados mostraram que é mais difícil produzir deformação mecânica em filmes de GaN dopado com Si e dopado com Mg do que no filme não dopado. A autorrecuperação que ocorre após a retirada da ponta foi estudada utilizando cristais de ZnO com diferentes orientações. O mecanismo de ativação térmica dos loops de discordância foi estudado através da observação da influência da temperatura no processo de autorrecuperação parcial dos cristais. Medidas de catodoluminescência foram usadas para identificar as distribuições de tensão associadas à deformação plástica permanente mostrando que esta induz regiões de tensão trativa ao longo das direções a 1120 nos filmes de GaN dopado e não dopado. / [en] In this work, the mechanical deformation of GaN films was studied by nanoindentation. A nanoindenter was used to induce the nucleation of mechanical defects on the samples surfaces in a controlled manner. The morphology of the indentations and the microstructure of the defects were studied using atomic force microscopy and transmission electron microscopy. The results showed that in the early stages of deformation, the nanoindentation process promotes slip at the atomic scale of the pyramidal planes of the crystal that can be reversed if the load is removed. If load is further increased, locking of these atomic plains occur leading to a hardened crystal region. It acts as an extension of the tip of the indenter redistributing the applied stress. At a critical stress, a major pop-in event occurs with the slip of the 1101, 1122 and 0001 plains leading then to irreversible plastic deformation. The influence of doping on the mechanical deformation has been studied and the results showed that it is more difficult to produce mechanical deformation in GaN films doped with Si and Mg doped than in undoped films. The self-recovery that occurs after removal of the tip was investigated using ZnO crystals with different orientations. The mechanism of thermal activation of dislocation loops was studied by observing the influence of temperature on the self-recovery process of the crystals. Cathodoluminescence measures were used to identify the resulting stress distributions associated with permanent plastic deformation showing that this induces tensile regions along the a 1120 directions in doped and undoped GaN films.

Polymer electrolyte membrane fuel cells (PEMFCs) have recently acquired much attention as alternatives to combustion engines for power conversion. The primary interest in fuel cell technology is the possibility of 60% power conversion efficiency as compared to the 30% maximum theoretical efficiency limited to combustion engines and turbines. Although originally conceived to work with hydrogen as a fuel, difficulties relating to hydrogen storage have prompted much effort in using other fuels. Small organic molecules such as alcohols and formic acid have shown promise as alternatives to hydrogen in PEMFCs due to their higher stability at ambient conditions. The drawbacks for using these fuels in PEMFCs are related to their incomplete oxidation mechanisms, which lead to the production of carbon monoxide (CO). When carbon monoxide is released in fuel cells it binds strongly to the platinum anode thus limiting the adsorption and subsequent oxidation of more fuel. In order to promote the complete oxidation of fuels and limit poisoning due to CO, various metal and metal oxide catalysts have been used. Motivated by promising results seen in fuel cell catalysis, this research project is focused on the design and fabrication of novel platinum-composite catalysts for the electrooxidation of methanol, ethanol and formic acid. Various Pt-composites were fabricated including Pt-Au, PtRu, Pt-Pd and Pt-CeO2 catalysts. Electrochemical techniques were used to determine the catalytic ability of each novel composite toward the electrooxidation of methanol, ethanol and formic acid. This study indicates that the novel composites all have higher catalytic ability than bare Pt electrodes. The increase in catalytic ability is mostly attributed to the increase in CO poison tolerance and promotion of the complete oxidation mechanism of methanol, ethanol and iv formic acid. Formulations including bi- and tri-composite catalysts were fabricated and in many cases show the highest catalytic oxidation, suggesting tertiary catalytic effects. The combination of bi-metallic composites with ceria also showed highly increased catalytic oxidation ability. The following dissertation expounds on the relationship between composite material and the electrooxidation of methanol, ethanol and formic acid. The full electrochemical and material characterization of each composite electrode is provided.

This dissertation addresses the fundamental problem of how to build computing systems for the nanoscale. With CMOS reaching fundamental limits, emerging nanomaterials such as semiconductor nanowires, carbon nanotubes, graphene etc. have been proposed as promising alternatives. However, nanoelectronics research has largely focused on a `device-first' mindset without adequately addressing system-level capabilities, challenges for integration and scalable assembly. In this dissertation, we propose to develop an integrated nano-fabric, (broadly defined as nanostructures/devices in conjunction with paradigms for assembly, inter-connection and circuit styles), as opposed to approaches that focus on MOSFET replacement devices as the ultimate goal. In the `fabric-centric' mindset, design choices at individual levels are made compatible with the fabric as a whole and minimize challenges for nanomanufacturing while achieving system-level benefits vs. scaled CMOS. We present semiconductor nanowire based nano-fabrics incorporating these fabric-centric principles called NASICs and N3ASICs and discuss how we have taken them from initial design to experimental prototype. Manufacturing challenges are mitigated through careful design choices at multiple levels of abstraction. Regular fabrics with limited customization mitigate overlay alignment requirements. Cross-nanowire FET devices and interconnect are assembled together as part of the uniform regular fabric without the need for arbitrary fine-grain interconnection at the nanoscale, routing or device sizing. Unconventional circuit styles are devised that are compatible with regular fabric layouts and eliminate the requirement for using complementary devices. Core fabric concepts are introduced and validated. Detailed analyses on device-circuit co-design and optimization, cascading, noise and parameter variation are presented. Benchmarking of nanowire processor designs vs. equivalent scaled 16nm CMOS shows up to 22X area, 30X power benefits at comparable performance, and with overlay precision that is achievable with present-day technology. Building on the extensive manufacturing-friendly fabric framework, we present recent experimental efforts and key milestones that have been attained towards realizing a proof-of-concept prototype at dimensions of 30nm and below.

The realm of the nano-world begins when things start getting smaller in size than one thousandth of the thickness of the human hair. Surface patterning at the nanoscale has started to find applications in information storage, self-cleaning of surfaces due to the "lotus effect", biocompatible materials based on surface roughness and many more. Several methods such as particle-beam writing, optical lithography, stamping and various kinds of self-assembly are widely used to serve the purpose of patterning smaller surface structures. However, globally much research is going into developing more efficient, reproducible and simple methods of patterning surfaces and in better controlling the order of these nanostructures. Researchers have always looked upon Nature to get inspiration and to mimic its model in engineering novel architectures. One of the methods used by this greatest artist (Nature) to make beautiful patterns around is through reaction diffusion based non-linear processes. Non-linear systems driven away from equilibrium sustain pattern only during the continuous dissipation of a regular flow of energy and are different from equilibrium processes that are converging towards a minimum in free energy (a. k. a. self-assembly). Dissipative pattern formation from micrometer to kilometers scale has been known but ordered patterns at nanoscale have never been achieved. In the process of thoroughly characterizing suitable substrates for nanoelectronics applications, we came across a remarkable process leading to the formation of highly ordered arrays of dimples on tantalum. The pattern formation happens in a narrow electrochemical windows which are functions of many parameters such as concentration, external applied voltage, temperature etc. After investigating the formation of dimples by performing spatio-temporal studies, we found that the underlying principles behind this unique way of engineering nano-structures have their roots in nonlinear interaction/reaction electro-hydrodynamics. We then have demonstrated the generality of this process by extending it to titanium, tungsten and zirconium surfaces. The pattern similar to Rayleigh-Bernard convection cells originates inside the electrochemical solution due to coupling among electrolyte ions during their migration across the electrochemical double layer (Helmholtz layer) and simultaneously imprints on the surface due to dissolution of metal oxide via etching. Based on these results we further postulate that, given appropriate electropolishing chemistry; these patterns can be formed on virtually any metal or semiconductor surface. The application of these nanostructures as nanobeakers for placing metal nanoparticles is also elucidated Highly porous materials such as mesoporous oxides are of technological interest for catalytic, sensing, optical and filtration applications: the mesoporous materials (with pores of size 2-50 nm) in the form of thin films can be used as membranes due large surface area. In the second part of this thesis, a new technique of making detachable ultrathin membranes of transition metal oxides is presented. The underlying concepts behind the detachment of membranes from the underlying substrate surface are discussed. The control on the size of the pores by modulating the voltage and concentration is also elucidated. The method is generalized by showing the similar detachment behavior on other metal oxide membranes.Thus, the results of this work introduces new techniques of engineering nanostructures on surfaces based on reaction-diffusion adaptive systems and contribute to the better understanding of electrochemical self-organization phenomena due to migration coupling induced electro-hydrodynamics. / Thesis / Doctor of Philosophy (PhD)

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.