Dissertations / Theses on the topic 'Microelectronics'

This thesis addresses the development of materials, technologies and circuits applied for the fabrication of a new class of microelectronic devices that are relying on a three-dimensional shape variation namely shapeable microelectronics. Shapeable microelectronics has a far-reachable future in foreseeable applications that are dealing with arbitrarily shaped geometries, revolutionizing the field of neuronal implants and interfaces, mechanical prosthetics and regenerative medicine in general. Shapeable microelectronics can deterministically interface and stimulate delicate biological tissue mechanically or electrically. Applied in flexible and printable devices shapeable microelectronics can provide novel functionalities with unmatched mechanical and electrical performance. For the purpose of shapeable microelectronics, novel materials based on metallic multilayers, photopatternable organic and metal-organic polymers were synthesized. Achieved polymeric platform, being mechanically adaptable, provides possibility of a gentle automatic attachment and subsequent release of active micro-scale devices. Equipped with integrated electronic the platform provides an interface to the neural tissue, confining neural fibers and, if necessary, guiding the regeneration of the tissue with a minimal impact. The self-assembly capability of the platform enables the high yield manufacture of three-dimensionally shaped devices that are relying on geometry/stress dependent physical effects that are evolving in magnetic materials including magentostriction and shape anisotropy. Developed arrays of giant magnetoimpedance sensors and cuff implants provide a possibility to address physiological processes locally or distantly via magnetic and electric fields that are generated deep inside the organism, providing unique real time health monitoring capabilities. Fabricated on a large scale shapeable magnetosensory systems and nanostructured materials demonstrate outstanding mechanical and electrical performance. The novel, shapeable form of electronics can revolutionize the field of mechanical prosthetics, wearable devices, medical aids and commercial devices by adding novel sensory functionalities, increasing their capabilities, reducing size and power consumption.

Glasses are used in microelectronic packaging for insulation and passivation purposes. To optimize the performance of these packages, it is necessary to investigate new glasses or improve on properties of the glasses in use. The insulating glass should have low dielectric constant, low dissipation factor, low glass transition temperature, high chemical resistivity, and a thermal expansion coefficient matching the substrate. In this study, various aluminoborophosphosilicate glasses containing Ca(Mg)O, Ca(Mg)F₂, and AlF₃ as flux were investigated. Processing temperatures for these glasses range from 1300°C to 1500°C. The coefficients of thermal expansion range from 4.52 μ/°C to 9.39 μ/°C. The dielectric constant as a function of frequency and composition is in the range of 4.1 to 5.2. The index of refraction for these glasses is in the range of 1.52 to 1.58. Glass transition and softening temperatures as low as 538°C and 622°C, respective, were found. Results of this investigation are discussed in terms of the possible use of aluminoborophosphosilicate glasses in microelectronic packaging.

As the size of hybrid microelectronics is reduced, the power density increases and thermal interaction between heat-producing devices becomes significant. A nondimensional model is developed to investigate the effects of heat source interaction on a substrate. The results predict the maximum temperature created by a device for a wide range of device sizes, substrate thicknesses, device spacings, and external boundary conditions. They can be used to assess thermal interaction for preliminary design and layout of power devices on hybrid substrates.

Previous work in this area typically deals with semi-infinite regions or finite regions with isothermal bases. In the present work, the substrate and all heat dissipating mechanisms below the substrate are modeled as two separate thermal resistances in series. The thermal resistance at the base of the substrate includes the bond to the heat sink, the heat sink, and convection to a cooling medium. Results show that including this external resistance in the model can significantly alter the heat flow path through the substrate and the spreading resistance of the substrate. Results also show an optimal thickness exists to minimize temperature rise when the Biot number is small and the device spacing is large.

Tables are presented which list nondimensional values for maximum temperature and spreading resistance over a wide range of substrate geometries, device sizes, and boundary conditions. A design example is included to demonstrate an application of the results to a practical problem. The design example also shows the error that can result from assuming an isothermal boundary at the bottom of the substrate rather than a finite thermal resistance below the substrate.

Several other models are developed and compared with the axisymmetric model. A one-dimensional model and two two-dimensional models are simpler than the axisymmetric model but prove to be inaccurate. The axisymmetric model is then compared with a full three-dimensional model for accuracy. The model proves to be accurate when sources are symmetrically spaced and when sources are asymmetrical under certain conditions. However, when the sources are asymmetrical the axisymmetric model does not always predict accurate results.

Master of Science

The dimensional scaling of CMOS technology is approaching its fundamental limits and alternate device architectures and more functional channel materials ensuring superior operation al sub l0mm length scales are required to realise post CMOS applications. Among the quest for alternate materials. carbon based nanostructures such as 1•0 carbon nanotubes and graphene nanoribbons are considered as one of the many possible candidates. Demonstration of these one dimensional graphitic prototypes in microelectronic industry are strongly hindered by several factors such as the difficulty in controlling band gap. precise positioning and manufacturing on wafer scale, controlling carrier type and carrier concentration. deposition of CMOS compatible gate di electric and formation of low resistive contacts. Amongst these. an absence of a finite band gap and issues in controlling carrier concentration in graphene presents a major challenge for this material to be considered in logic devices. Moreover. both carbon nanotubes and graphene nanoribbons exhibit a diameter/width dependent band gap. At the nanoscale however. estimating the band gap both from experiment and theory is fraught with difficulties related to the underlying assumptions. Hence an accurate theoretical prediction of the performance of these material in future microelectronic devices is essential at this point since the International Technology Road map for Semiconductors (ITRS) -2011 edition strongly recommends carbon based nanostructures as an alternate channel material al for post CMOS logic applications. Also the technology requires novel band gap engineering techniques and doping g strategies without destruction of the intrinsic properties to enable application of graphene in various sectors such as microelectronics. photonics, photovoltaics and biosensors. In this thesis. suitable device geometries of such carbon based materials and their performance evaluation using realistic band gap values is investigated. Zig-zag nanotubes in the diameter range 0.55 - 1.26nm are considered in the study. Both MOSFET and tunnel FET device geometrics arc considered, but more focus is given To the tunnel-FET considering its energy efficient operation. The evaluation is made using non-equilibrium Greens function based numerical simulation with band gap values which have been calculated from the slate of an many-body perturbation theory GW method by P.Umari el al. Comparison of on-off ratio. device delay and saturation current in nanotube tunnel FETs reveals a considerable difference from previous evaluations made in the literature. Analysis reveals that. among the semiconducting zig-zag chiralities considered here, (11,0) nanotube devices in the tunnel FET configuration exhibits the best on-off ratio. device delay and saturation current which meets the ITRS-20Il requirement of low-operating power technology for 2020. Comparison of nanotube tunnel FETs with graphene nanoribbon tunnel FETs reveals that nanotube FETs deliver high on-off ratio and saturation current and exhibit comparable device delay. Among the considered families of nanoribbons, a l6AGNR exhibits best on -off ratio and device delay which operates within the future ITRS requirement. An atomic level investigation of the fundamental properties of a new type of metalgraphene system fanned by intercalation of gold atoms in epitaxial graphene is also presented. The effects of gold deposition on monolayer graphene (MG) epitaxied on SiC (0001) substrate are examined via Scanning Tunneling Microscopy (STM) and Spectroscopy (STS). Gold atoms exhibit mainly two types of self- assembly process below the graphene layer and the resulting gold-graphene systems exhibits contrasting electronic properties from each other. Insertion of a monolayer of gold below the monolayer graphene opens up a 100meV band gap in the electronic spectrum of graphene and creates a finite p-doping in graphene. While gold atoms intercalated in the form of atomic clusters shows negligible doping effect. Finally, signatures of a superlattice structure composed of a quasiperiodic arrangement of atomic gold clusters below an epitaxied graphene layer are examined Using dispersive Raman spectroscopy. The gold- graphene system exhibits a laser excitation energy dependant red shift of the 2D mode as compared to pristine epitaxial graphene. The phonon dispersion in pristine and gold intercalated graphene are mapped using experimentally observed Raman signatures and third-nearest neighbour (3NN) tight binding band structure model. This reveals that the observed behaviour is caused by modifications of the phonon. dispersion rather than changes in electronic structure. The intercalated gold atoms are found to restore the phonon band structure of epitaxial graphene towards that of free standing graphene.

Thesis (M. S.)--Mechanical Engineering, Georgia Institute of Technology, 2010.
Committee Chair: Graham, Sam; Committee Member: Joshi, Yogendra; Committee Member: Kalaitzidou, Kyriaki. Part of the SMARTech Electronic Thesis and Dissertation Collection.

The increasing demand for effective and compact thermal solutions for the next generation of thin and high-power density consumer electronics is challenging the capability of miniature mechanical systems to meet the required cooling performance. Due to their attractive and unique advantages with no moving parts, design flexibility, small-scale structure, low height profile, silent operation, and effective flow generation, electrohydrodynamic (EHD) air movers are well positioned to become a key emerging cooling technology as alternative to conventional rotary fans. In its general objective, this thesis aims to investigate the benefits and highlight the features of EHD air movers as a thermal management cooling solution in advanced and small-scale microelectronics, supporting all previous efforts in this direction. Due to the strong influence of the geometric parameters of EHD devices on the corona discharge process and the resulting EHD flow, numerical modelling represents a powerful tool to design and optimize EHD devices, especially of complex and small-scale structures, where the capability of experimental investigations is limited or challenging. This study presents an accurate and validated numerical method to solve the coupled equations of electrostatics, charge transport and fluid flow for the two-dimensional (2D) modelling of EHD airflow induced through a wire-to-plane/grid channel configuration, and is the first to develop a three-dimensional model (3D) that couples the EHD flows with conjugate heat transfer modelling. Based on thermal management requirements and from a design perspective, a comprehensive investigation and analysis into the influence of geometric parameters on the efficiency of EHD wire-to-grid blowers is performed and optimal configurations are proposed for a range of heights from 9 to 15 mm. Results reveal that using fine emitter wires is more efficient than thicker ones, and the grounded electrode locations affect significantly the electric field distribution and the blower efficiency. It is also found that using the grid as a further collector increases the blower performance, with higher flow production, lower operating voltage and reduced blower size. Further numerical developments are devoted to optimize the configuration of miniature wire-to-plane EHD blowers for heights up to 10 mm, which is the most preferred geometry for integration in the cooling systems of thin electronic applications. For ranges of fixed operating power and voltage, the efficient optimized electrode gaps are predicted and defined by simple expressions. The influence of channel sidewall on the EHD flow rate and velocity profile are investigated and the results show that the 2D modelling is valid to effectively predict flow rates produced by wide and short EHD blowers compared to that obtained by 3D simulations. A combined EHD air blower that enables a reduction in the level of applied voltage and a control of flow production is developed. Performance comparisons against commercial rotary blowers demonstrate that the optimized miniature EHD blowers are more competitive for cooling miniaturized and extended heated surfaces based on blower size, flow rate with uniform velocity profile, and power consumption. A novel design of an EHD system integrated with compact heat sinks is presented as a thermal management cooling solution for advanced and thin consumer applications. Results of a parametric study demonstrate that the EHD system offers flexible structure design with the ability to reduce the height and increase the width as required, providing a unique feature to be installed in low-profile laptops. Moreover, compared to traditional cooling systems used in the current standard low power laptops, the proposed EHD system offers promising cooling performance with higher thermal design power (TDP), reduced thermal solution volume and lower height profile.

A neural implant giving an amputee a sense of feeling back in their prosthetic limb could help millions of people live happier, more productive lives. Tactile feedback is commonly targeted, however, it is the lesser known sense of proprioception that is crucial for smooth, coordinated limb control and non-visual limb awareness - both of which are high priorities for amputees. This thesis describes research carried out to progress the development and creation of aproprioceptive neural prosthesis targeted at trans-humeral upper limb amputees. Firstly a review of proprioceptive neural prosthesis design considerations and challenges is presented. The purpose of which is to identify areas requiring further development and to identify a prototype target system that focuses and scopes design effort. Then 3 technical chapters cover research into: (1) Combining efficient implementations of biomechanical and proprioceptor models in order to generate signals that mimic human muscular proprioceptive patterns. A neuromusculoskeletal model of the upper limb with 7 degrees of freedom and 17 muscles is presented and generates real time estimates of muscle spindle and Golgi Tendon Organ neural firing patterns. (2) An 8 channel energy-efficient neural stimulator for generating charge-balanced asymmetric pulses. Power consumption is reduced by implementing a fully-integrated DC-DC converter that uses a reconfigurable switched capacitor topology to provide 4 output voltages for Dynamic Voltage Scaling (DVS). A novel charge balancing method is implemented which has a low level of accuracy on a single pulse and a much higher accuracy over a series of pulses. The method used is robust to process and component variation and does not require any initial or ongoing calibration. (3) A non-invasive proprioceptive feedback trial platform (using vibration induced proprioception) for testing modelled neural signals. A low cost vibration device is designed and tested, identifying key issues with this form of non-invasive feedback.

Aerosol-based direct-write refers to the additive process of printing CAD/CAM features from an apparatus which creates a liquid or solid aerosol beam. Direct-write technologies are poised to become useful tools in the microelectronics industry for rapid prototyping of components such as interconnects, sensors and thin film transistors (TFTs), with new applications for aerosol direct-write being rapidly conceived. This research aims to review direct-write technologies, with an emphasis on aerosol based systems. The different currently available state-of-the-art systems such as Aerosol Jet? CAB-DW?, MCS and aerodynamic lenses are described. A review and analysis of the physics behind the fluid-particle interactions including Stokes and Saffman force, experimental observations and how a full understanding of theory and experiments can lead to new technology such as nozzle designs are presented. Finally, the applications of aerosol direct-write for microelectronics are discussed in detail including the printing of RFID antennas, contacts and active material for TFTs, the top metallization layer for solar cells, and interconnects for circuitry.

Electromigration (EM) is a phenomenon that occurs in metal conductor carrying high density electric current. EM causes voids and hillocks that may lead to open or short circuits in electronic devices. Avoiding these failures therefore is a major challenge in semiconductor device and packaging design and manufacturing, and it will become an even greater challenge for the semiconductor assembly and packaging industry as electronics components and interconnects get smaller and smaller. According to the assembly and packaging section of the International Technology Roadmap for Semiconductor (ITRS) developed in 2007 and 2009 [1] [2], EM was a near term threat for the interconnecting part of semiconductor, devices and packaging methods such as flip chip, and Ball Grid Array (BGA). In the industry, EM-aware designs are mainly based on design rules that are derived from empirical laws which do not help understand complicated EM processes and therefore can’t be used to carry out accurate predictions for EM failures of sophisticated components in varied environmental conditions. In this work, novel numerical modelling methods of EM in micro-electronics devices have been developed and the methods have been used to analyse EM process in a lead free solder thin film, and to optimize the design of electronic components in order to reduce the risk of EM relative failure. EM is an atomic diffusion process that is driven by a high density electric current, but it is strongly affected by temperature and its gradient as well as stress distribution. In order to model EM accurately, the interacting electrical, thermal, and mechanical phenomena must all be solved simultaneously. In this work, a novel multi-physics modelling method has been proposed and developed to include all of the above mentioned physical phenomena using unstructured Finite Volume (FV) and Finite Element (FE) techniques. The methods have been implemented on the multi-physics software package PHYSICA. Comparing with existing methods, this fully coupled solution method is a significant improvement that will facilitate further development of electronics design and optimization tool as well as new research work that helps understand EM phenomenon. The developed models can be used to simulate the whole process of EM, predict voids initiation lifetime of electronics products or test specimens. In today’s electronics manufacturing, lead-free solder alloys are used as interconnect. As in copper or aluminium interconnect EM has become a threat to device reliability as current density increase in solder joints with diminishing sizes. In this work, computer simulation methods have been used to analyse the experimentally observed EM process in a thin film solder. The experiment was designed in such a way that effects of temperature and stress gradients can be avoided. The advantage of this experimental method is that the electric current effect is isolated which makes analysis and model validation easier. In this work, the predicted voids locations are consistent with experimental results. In this work, numerical examples are given to illustrate how interconnect designs can be made more EM failure resistant. The ultimate aim of the research is to understand EM and to develop techniques that predict EM accurately so that EM-aware designs can be made easier.

Thesis (M.S.)--State University of New York at Binghamton, Thomas J. Watson School of Engineering and Applied Science, Department of Systems Science and Industrial Engineering, 2007.
Includes bibliographical references.

Thesis (M. Phil.)--Hong Kong University of Science and Technology, 2004.
Includes bibliographical references (leaves 116-127). Also available in electronic version. Access restricted to campus users.

The objective of this research is to develop a design framework for virtual prototyping of electronic packaging. This framework couples computational mechanics and fluid dynamics, based on finite volume method with integrated finite element routines, with numerical optimisation and statistical methods. This integrated approach is intended as a modelling tool for calculating optimal design solutions for electronic packaging and component assembly with a focus on die reliability and the thermal management. The motivation is to introduce numerical optimisation theory as an approach for a fast, systematic and automated design approach for wide range microelectronics applications. The proposed methodology will also benefit from multi-physics numerical analysis to predict complex behaviour of electronic packages, systems and processes subject to different operational or environmental conditions. This thesis demonstrates multi-physics modelling (i.e. integrated solutions for fluid flow, heat transfer and stress) coupled with gradient/non-gradient based numerical optimisation techniques and associated statistical methods. An explanation and comparison of the two approaches to numerical optimisation — (1) Response Surface Methodology (RSM) based on Design of Experiments (DoE) and (2) direct gradient based and non-gradient methods - are given. Both the advantages and limitations of these virtual design strategies, with respect to their integration with multi-physics modelling, are discussed and demonstrated. This integrated multiphysics/optimisation design approach is demonstrated on a variety of problems from the area of microelectronics design and packaging. The thesis demonstrates this for three industrial examples. These are: The software packages used to develop the design tool and to undertake the outlined studies are PHYSICA and VisualDOC. PHYSICA is a multiphysics finite volume based simulation tool with integrated modules for finite element solid mechanics analysis. The software framework is detailed in Chapter 2, Section 2.4 and further in Chapter 4. The VisualDOC tool offers a collection of numerical optimisation routines and modules for statistical analysis (Design of Experiments) and approximate Response Surface modelling. VisualDOC framework is discussed in Chapter 4, Section 4.8.

This research analyzes the re-oxidation annealing process of Barium titanate thin films on copper foils made by Chemical Solution Deposition. During this anneal, the temperature and oxygen pressure settings must be optimized to ensure the elimination of oxygen vacancies without oxidizing the copper foil substrate. This research utilizes Design of Experiments (DOE) to study the impact of re-oxidation furnace temperature and pressure on the dielectric loss tangent response. Two designs of experiments were run. The first experiment, a 32 DOE, examined a large range of temperature and pressure levels. Due to the high susceptibility of uncontrollable factors such as humidity and film position in the crystallization anneal furnace, an adequate model could not be developed. However, the temperature at 550ºC and a pressure of 10-5 Torr yielded a lower mean and standard deviation of the loss tangent response. A second and smaller scale experiment, a 22 with a center point, was run around 550ºC and 10-5 Torr to determine if more optimal temperature and pressure settings existed in the local area. Two second order response surface models were developed from two crystallization anneals that were statistically significant. The most significant finding was that the optimum level for temperature and pressure in the re-oxidation anneal furnace in this experiment is 550ºC and 2x10-5 Torr. While the models concluded that the temperature, pressure, temperature quadratic, and interaction between pressure and temperature were important effects in the model, there were differences in the curvature of the models due to the temperature quadratic effect.

Thesis (M. Eng.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 2006.
Includes bibliographical references (p. 57-58).
We developed the MIT Integrated Microelectronics Device Experimentation and Simulation iLab, a new online laboratory that combines and significantly upgrades the capabilities of two existing online microelectronics labs: WebLab, a device characterization lab, and WebLabSim, a device simulation lab. The new integrated tool allows users to simultaneously run experiments on actual devices and simulations on the virtual ones, as well as to compare the results of the two. In order to achieve this, we considerably extended the capabilities of the original clients. We added the ability to graph the results of multiple experiments and simulations simultaneously, on top of each other, which allows for much easier comparison. We also added the ability to load, view and graph the results of experiments and simulations that were ran at any point in the past, even when the corresponding lab configurations are no longer available. Our hope is that this new integrated iLab will enrich microelectronics teaching and learning by allowing students to compare real life device behavior with theoretical expectations.
by Boris Cukalovic.
M.Eng.

Doutoramento em Ciência e Engenharia de Materiais
Devido à redução de dimensões e ao aumento da velocidade de processamento de dados nos dispositivos microeletrónicos baseados em semicondutores convencionais, estão a ser exploradas abordagens inovadoras envolvendo novos materiais tais como óxidos funcionais. Com o rápido desenvolvimento da indústria eletrónica existe uma maior necessidade de elevado desempenho, de elevada fiabilidade, e de componentes eletrónicos miniaturizados integrados em vários dispositivos. A fim de tornar os dispositivos amplamente acessíveis e de fácil utilização, requisitos adicionais devem ser considerados: o tamanho e peso desejados, o custo reduzido, o baixo consumo de energia e a portabilidade. Materiais funcionais de baixa dimensionalidade são muito promissores para cumprir essas exigências. Em particular, os ferroeléctricos de filmes finos bidimensionais (2D) têm recebido grande atenção devido à sua crescente utilização em memórias não voláteis, detectores piroelétricos, transdutores piezoeléctricos miniaturizados e dispositivos sintonizáveis de micro-ondas. A temperatura de cristalização é um parâmetro chave na preparação de ferroelétricos 2D. Muitos filmes finos ferroelétricos são cristalizados a temperaturas >600 °C. Esses valores estão acima da temperatura que certos elementos do dispositivo funcional podem suportar. Recentemente, este facto tornou-se ainda mais importante, devido às promissoras aplicações que podem ser consideradas caso os ferroeléctricos 2D sejam compatíveis com substratos poliméricos flexíveis de baixo custo e de baixo ponto de fusão. A compatibilidade de filmes finos ferróicos com estes últimos tipos de substratos é muito difícil, mas se conseguida pode ampliar acentuadamente a gama de aplicações para os mais recentes requisitos de eletrónica flexível e microeletrónica, onde dispositivos leves e baratos são exigidos. Neste trabalho, é implementada uma combinação da modificação da química de precursores e assistência por luz UV, com promoção simultânea da cristalização pela introdução de sementes nanocristalinas na solução precursora, para a fabricação de filmes finos ferróicos sem chumbo - Método de Precursores Fotossensíveis Semeados. Neste contexto, o principal objetivo deste trabalho foi fabricar filmes finos sem chumbo BiFeO3 (BFO) e Na0.5Bi0.5TiO3 (NBT) a baixas temperaturas (~300 °C) com uma resposta ferroelétrica competitiva. Além disso, a investigação do efeito do elétrodo-base sobre as propriedades dielétricas e ferroelétricas de filmes finos de BFO foi levada a cabo, e a comparação entre o comportamento de condensadores de BFO com base em IrO2, LaNiO3 (LNO) e Pt foi estabelecida. Adicionalmente, os efeitos dos vários eléctrodos sobre a microestrutura de filmes finos ferroeléctricos de BFO foram estudados por microscopia eletrónica de transmissão (TEM) de alta resolução. Primeiramente, filmes finos finos de perovesquite BFO e NBT foram preparados sobre substratos de silício revestidos com Pt, por deposição de solução química. Os filmes finos de BFO foram preparados a temperaturas na gama de 400-500 °C, a partir de soluções de precursores estequiométricas e com excesso de Bi. Os filmes de BFO cristalinos foram obtidos a 400 °C, o limite inferior de temperatura. Os filmes preparadas com excesso de Bi possuem curvas de histerese ferroelétrica mais definidas do que aqueles sem qualquer excesso, para filmes com espessuras ~150 nm. Uma vez que as densidades de corrente de fuga nos filmes finos diminuem com a diminuição da temperatura de processamento, a polarização de filmes finos de BFO preparados com excesso Bi e recozidos a 400 e 450 °C pode ser efetivamente comutada à temperatura ambiente. Obtiveram-se valores de polarização remanescente de Pr ~10 e ~60 μC/cm2 com campos coercivos de EC ~ 205 e 235 kV/cm para os filmes finos preparados a 400 e 450 °C, respectivamente. Os filmes finos de NBT foram preparados a temperaturas entre 400 e 650 °C. As propriedades estruturais e ferroelétricas dos filmes foram examinadas. A constante dieléctrica observada e as perdas dieléctricas a 100 kHz são 616 e 0,032, respectivamente, enquanto que a polarização remanescente observada e o campo coercivo são Pr ~ 24 μC/cm2 e EC ~ 215 kV/cm, respectivamente para o filme de NBT recozido a 650 °C. O recozimento térmico, em atmosfera de oxigénio após cada camada de revestimento, é eficaz na promoção da cristalização do filme na fase de perovesquite romboédrica a uma baixa temperatura de 400 °C. No entanto, obteve-se um ciclo P-E quase linear para os filmes NBT cristalizados a 400 °C devido à sua incipiente cristalinidade. Os filmes finos de BFO foram depositados numa gama de elétrodos para determinar o seu papel no controlo da formação de fases e da microestrutura. A cristalização em elétrodos de óxido seguiu a sequência: amorfa → Bi2O2(CO3) → perovesquite, enquanto que nos elétrodos de Pt cristalizaram diretamente a partir da fase amorfa. Os elétrodos de IrO2 promoveram a formação da fase de perovesquite à temperatura mais baixa e o LNO induziu adicionalmente o crescimento epitaxial local. O LNO tem a estrutura de perovesquite com o parâmetro de rede a = 0.384 nm, compatível com o de BFO, a = 0.396 nm, e assim a epitaxia é mais provável. Todas as composições exibiram precipitados inteiramente coerentes ricos em Fe dentro do interior de grão da matriz de perovesquite, enquanto que a incoerente segunda fase de Bi2Fe4O9 foi também observada nos limites de grão de BFO crescido em eléctrodos de Pt. Esta última pode ser observada por difração de raios X, bem como TEM, mas os precipitados coerentes foram observados apenas por TEM, principalmente evidenciados pelo seu contraste Z em imagens de campo escuro anular. Estes dados têm consequências acentuadas permitindo alargar a utilização de filmes de BFO sob campo aplicado, a aplicações como atuadores, sensores e aplicações de memória. Em seguida, os filmes finos de BFO foram depositados em substratos de Si com elétrodos distintos, como Pt, LNO e IrO2, para investigar o efeito do elétrodo-base sobre o crescimento e as propriedades elétricas do BFO. Todas os filmes de BFO são compostos por grãos colunares cujo tamanho é dependente do elétrodo-base. Não se observou textura para filmes de 320 nm de espessura fabricados em Pt orientado (111). Os filmes sobre eléctrodos de óxido, em particular sobre LNO são altamente orientados no plano (012). A grande polarização remanescente em BFO/Pt e BFO/IrO2 é atribuída à alta contribuição de corrente de fuga. Os filmes BFO de 400 nm de espessura em LNO possuem uma baixa densidade de corrente de fuga ~4 × 10-6 A/cm2, uma grande polarização remanescente de 50 μC/cm2 e um pequeno campo coercitivo de 180 kV/cm à temperatura ambiente. Demonstramos que as camadas de LNO aumentam a cristalinidade e a orientação de filmes finos BFO, o que se reflete nas suas propriedades funcionais. Este estudo mostra que, além da simples necessidade de filmes monofásicos, os elétrodos de óxido de metal têm um impacto relevante no desenvolvimento de filmes finos BFO de alta qualidade fabricados por métodos químicos de deposição de solução. Estes resultados têm uma implicação grande para a fabricação de dispositivos BFO baseados em filmes finos. Finalmente, provamos que é possível fabricar diretamente filmes finos de BFO sem chumbo em substratos flexíveis de poliamida com funcionalidades ferroelétricas e magnéticas (multiferroicidade) à temperatura ambiente. O nosso método inovador, baseado em soluções de Precursores Fotossensíveis e nanosementes cristalinas, foi usado com sucesso para diminuir a temperatura de cristalização de filmes finos de BFO até uma temperatura tão baixa quanto 300 °C, a mais baixa temperatura reportada até agora para a preparação de filmes finos multiferróicos de BFO. Apesar deste excepcionalmente baixo nível térmico, obtém-se uma polarização remanescente Pr de 2.8 μC/cm2 para os filmes semeados + UV, com um campo coercitivo EC de 300 kV/cm. A estratégia de síntese baseada na utilização de precursores fotossensíveis sementados pode ser transferida para qualquer outra família de óxidos metálicos funcionais.
With the dimensions reduction and data processing speeds increasing of conventional semiconductor based microelectronic devices, innovative approaches involving new materials such as functional oxides are being explored. With the rapid development of the electronics industry there is a need for high performance, high reliability and miniaturized electronic components integrated into various devices. In order to make the devices user friendly and widely accessible, additional requirements should be considered: the desired size and weight, low cost, low power consumption, and portability in addition to high levels of functionality. Low dimensional functional materials hold great promises to fulfil those requirements. In particular, two-dimensional (2D) thin film ferroelectrics have received wide attention because of their growing use as non-volatile memories, pyroelectric detectors, miniaturized piezoelectric transducers and tunable microwave devices. Crystallization temperature is a key parameter in preparation of 2D-ferroelectrics. Many ferroelectric thin films are crystallized at temperatures >600 °C. This is above the temperature that certain elements of the functional device can withstand. Recently it became even more important due to promising applications that can be envisaged if 2D-ferroelectrics will be compatible with low cost, low melting temperature flexible polymeric substrates. The compatibility of ferroic thin films with those last types of substrates can markedly widen the range of applications towards the most recent requirements of flexible electronics and microelectronics, where lightweight and cheap devices are demanded. In this work, a combination of the modification of precursor chemistry and the assistance of UV-light, with simultaneous promotion of crystallization by introducing nanocrystalline seeds in the precursor solution, is implemented to fabricate lead-free ferroic thin films - Seeded Photosensitive Precursor Method. Within this context, the main objective of this work was to fabricate lead-free BiFeO3 (BFO) and Na0.5Bi0.5TiO3 (NBT) thin films with a competitive ferroelectric response at low temperatures. Moreover, investigations of the effect of the bottom electrode on the dielectric and ferroelectric properties of BFO thin films was conducted and the comparison between the behavior of IrO2, LaNiO3 (LNO) and Pt based BFO capacitors established. Additionally, the effects of these various bottom electrodes on the microstructure of BiFeO3 ferroelectric films was studied by high-resolution TEM. Firstly, BFO and NBT perovskite thin films were prepared on Pt-coated silicon substrates by chemical solution deposition. BFO was prepared at temperatures in the range 400-500 °C, and from stoichiometric and Bi excess precursor solutions. Crystalline BFO films were obtained at the lowest temperature limit of 400 °C. The films prepared with Bi excess possess more defined ferroelectric hysteresis loops than those without any excess; for films with thicknesses ~150 nm. As the leakage current densities in the films decrease with decreasing the processing temperature, polarization of BFO films prepared with Bi excess and annealed at 400 and 450 °C can be effectively switched at room temperature. Remanent polarization values of Pr ~ 10 and ~60 μC/cm2 with coercive fields of EC ~ 205 and 235 kV/cm were obtained for the films prepared at 400 and 450 °C, respectively. NBT thin films were prepared at temperatures from 400 to 650 °C. Structural and ferroelectric properties of the films were examined. The observed dielectric constant and dielectric losses at 100 kHz are 616 and 0.032, respectively, while the observed remanent polarization and coercive field are Pr ~ 24 μC/cm2 and EC ~ 215 kV/cm, respectively for the NBT film annealed at 650 °C. Thermal annealing in an oxygen atmosphere after each layer of coating is effective in promoting crystallization of the film into rhombohedral perovskite phase at a low temperature of 400 °C. However, almost linear, P-E loop was obtained for those NBT films crystallized at 400 °C due to incipient crystallinity. BFO thin films were grown on a range of electrodes to determine their role in controlling phase formation and microstructure. The crystallization on oxide electrodes followed the sequence: amorphous → Bi2O2(CO3) → perovskite, while those on Pt crystallized directly from the amorphous phase. IrO2 electrodes promoted perovskite phase formation at the lowest temperature and LaNiO3 additionally induced local epitaxial growth. LNO has the perovskite structure with lattice parameter a = 0.384 nm, compatible with that of BFO, a = 0.396 nm and thus epitaxy is more likely. It was observed for the first time that all compositions exhibited fully coherent Fe-rich precipitates within the grain interior of the perovskite matrix, whereas incoherent Bi2Fe4O9 second phase was also observed at the grain boundaries of BFO grown on Pt electrodes. The latter could be observed by X-ray diffraction as well as transmission electron microscopy (TEM) but coherent precipitates were only observed by TEM, principally evidenced by their Z contrast in annular dark field images. These data have pronounced consequences for the extended use of BFO films under applied field for actuator, sensor and memory applications. Then, BFO thin films were deposited on Si-based substrates with distinct electrodes, such as Pt, LNO, and IrO2, in order to investigate the effect of bottom electrode on the growth and electrical properties of BFO. All BFO films are composed of columnar grains which size is dependent on the bottom electrode. No texture was observed for 320 nm thick films fabricated on (111) oriented Pt. Films on oxide electrodes, in particular on LNO are highly (012) oriented. The large remanent polarization in BFO/Pt and BFO/IrO2 is attributed to the high leakage current contribution. 400 nm thick BFO films on LNO possess a low leakage current density ~4 × 10-6 A/cm2, a large remanent polarization of 50 μC/cm2 and a small coercive field of 180 kV/cm at room temperature. We demonstrate that LNO layers enhance the crystallinity and orientation of BFO thin films, which is reflected in their functional properties. This study shows that besides the simple need of monophasic films metal oxide electrodes have a relevant impact on the development of high quality BFO thin films fabricated by chemical solution deposition methods. These results have a broad implication for the fabrication of BFO thin film based devices. Finally, we prove that it is possible to directly fabricate lead-free BFO thin films on flexible polyamide substrates with ferroelectric and magnetic functionalites (multiferroicity) at room temperature. Our own proprietary novel solution-based Seeded Photosensitive Precursor Method was successfully used to decrease the crystallization temperature of BFO thin films down to a temperature as low as 300 °C, the lowest reported up to now for the preparation of multiferroic BFO thin films. Despite this exceptionally low thermal budget a remanent polarization Pr of 2.8 μC/cm2 is obtained for the seeded + UV films, with a coercive field EC of 300 kV/cm. The synthesis strategy based on the use of seeded photosensitive precursors can be transferred to any family of functional metal oxide.

Thesis (M. Phil.)--Hong Kong University of Science and Technology, 2004.
Includes bibliographical references (leaves 60-68). Also available in electronic version. Access restricted to campus users.

Thesis (Ph.D)--Mechanical Engineering, Georgia Institute of Technology, 2010.
Committee Chair: Samuel Graham; Committee Member: Donald Dorsey; Committee Member: Douglas Yoder; Committee Member: Michael Leamy; Committee Member: Sankar Nair; Committee Member: Zhuomin Zhang. Part of the SMARTech Electronic Thesis and Dissertation Collection.

Owing to their many interesting characteristics, the application of metal oxide based electronics has been growing at a considerable rate for the past ten years. High performance, optical transparency, chemical stability and suitability toward low cost deposition methods make them well suited to a number of new and interesting application areas which conventional materials such as silicon, or more recently organic materials, are unable to satisfy. The work presented in this thesis is focussed on the optimisation of high performance metal oxide based electronics combined with use of spray pyrolysis, as a low cost deposition method. The findings presented here are split into three main areas, starting with an initial discussion on the physical and electronic properties of films deposited by spray pyrolysis. The results demonstrate a number of deposition criteria that aid in the optimisation and fabrication of high performance zinc oxide (ZnO) based thin-film transistors (TFTs) with charge carrier mobilities as high a 20 cm2/Vs. Solution processed gallium oxide TFTs with charge carrier mobilities of ~0.5 cm2/Vs are also demonstrated, highlighting the flexibility of the deposition method. The second part of the work explores the use of facile chemical doping methods suitable for spray pyrolysed ZnO based TFTs. By blending different precursor materials in solution prior to deposition, it has been possible to adjust certain material characteristics, and in turn device performance. Through the addition of lithium it has been possible alter the films grain structure, leading to significantly improved charge carrier mobilities as high as ~54 cm2/Vs. Additionally the inclusion of beryllium during film deposition has been demonstrated to control TFT threshold voltages, leading to improved integrated circuit performance. The final segment of work demonstrates the flexibility of spray pyrolysis through the deposition of a number of high-k dielectric materials. These high performance dielectrics are integrated into the fabrication of TFTs already benefiting from the findings of the previously discussed work, leading to highly optimised low-voltage TFTs. The performance of these devices represent some of best currently available from solution processed ZnO TFTs with charge carrier mobilities as high as 85 cm2/Vs operating at 3.5 V.

Doutoramento em Ciência e Engenharia de Materiais
Nos dias que correm os engenheiros de circuítos na área da microelectrónica são confrontados com a necessidade de desenhar circuítos que permitam um maior transporte de informação numa menor largura de banda, que consumam menos energia e ao mesmo tempo com a necessidade de criar produtos de dimensões menores e mais flexíveis, com maiores níveis de integração, operação a frequências mais elevadas e a custos mais reduzidos. Neste contexto, a substituição de componentes dieléctricos cerâmicos na forma de monolitos, que são parte integrante de determinados dispositivos microelectrónicos que operam a frequências elevadas (filtros e antennas, por exemplo), por dieléctricos processados na forma de filmes espessos está sob consideração. Com esta aproximação espera-se, por um lado conseguir uma redução do tamanho do dispositivo e dos custos associados à sua produção e por outro lado, e de particular relevância, explorar as oportunidades criadas pela possibilidade de processar filmes conformes com substratos de diferentes formas e de natureza metálica. Novas estruturas e concepções para dispositivos que operam a frequências elevadas deverão ser criadas. Ao mesmo tempo, crê-se contribuir para o desenvolvimento de processos de fabrico de produção em massa de filmes de materiais dieléctricos com desempenho reproductível e a baixos custos. As técnicas de preparação de filmes finos incluem a fabricação por cinta (“tape casting”), por impressão em tela (“screen printing”), por jacto de tinta (“jet printing”) e por deposição electroforética (“electrophoretic deposition”, EPD). A importância da deposição electroforética advém das suas características únicas, que incluem a simplicidade e flexibilidade na aplicação a vários tipos de materiais e combinações de materiais numa gama alargada de formas e dimensões de estruturas, na relação fabricação – custos e na capacidade de dimensionar o processo à escala industrial para fabricação de volumes elevados de produtos e de grandes dimensões. Concomitantemente, quando comparado com os outros processos de fabricação de filmes espessos, a deposição electroforética permite a produção de camadas de uniformidade excepcional com um fácil controlo da sua espessura. Desta forma a fabricação de filmes espessos por deposição electroforética vai de encontro às actuais necessidades da indústria da microelectrónica no que respeita à substituição dos componentes dieléctricos monolitos em uso hoje em dia. Em relação aos materiais, os dieléctricos que podem ser utilizados como componentes de dispositivos de operação às frequências das microondas, deverão possuir perdas dieléctricas baixas, factor de qualidade (definido como o inverso das perdas dieléctricas) elevado, permitividade dieléctrica elevada e coeficiente de temperatura da permitividade baixo. De entre os dieléctricos com baixas perdas, o sistema BaO-Nd2O3-TiO2 representa uma importante família comercial de materiais para utilização às frequências das microondas, em particular a composição 1:1:5, que, porque reúne as características acima mencionadas, é um material de referência para estas aplicações. Embora cerâmicos de BaO-Nd2O3-TiO2 estejam actualmente em produção e comercialmente disponíveis em resonadores, filtros e substratos, entre outras aplicações, o uso de filmes espessos de BNT não foi até à realização deste estudo, referido. Neste trabalho, é explorada a fabricação por deposição electroforética de filmes espessos de BaNd2Ti5O14 (BNT). Para tal é conduzido um estudo sistemático do processo de deposição, desde a prova de conceito de aplicação do processo de deposição electroforética até à definição de condições reprodutíveis e optimizadas de deposição. São utilizados pós comerciais e pós fabricados em laboratório que foram depositados sobre folhas metálicas flexíveis de platina e substratos de alumina. Inicia-se o trabalho pela prova do conceito de aplicação da deposição electroforética ao fabrico de filmes de BNT. Filmes de BNT com 12 a 52 μm de espessura são fabricados a partir de pós de BNT comerciais sobre substratos de platina. Para melhorar a densidade em verde e a microestutura dos filmes obtidos recorre-se a uma etapa intermédia de prensagem isostática dos filmes em verde. O efeito da espessura dos filmes nas propriedades dieléctricas a baixas frequências é analisado. À medida que a espessura do filme aumenta, as propriedades dieléctricas dos filmes de BNT aproximam-se das propriedades dos cerâmicos de BNT em termos de permitividade e perdas dieléctricas. Filmes de BNT com 52 μm de espessura e sinterizados a 1300 °C durante 1 h exibem uma constante dieléctrica e uma perda dieléctrica de 107 e 0.0006 ou um factor de perda Q de 1600 a 1 MHz, respectivamente. A variação de permitividade dieléctrica é inferior a 0.02 % a um campo eléctrico de ±8 kV/cm e na gama de temperatura entre 30-120 °C é abaixo de +58.5 ppm/°C. O estudo revela ainda que não há degradação das propriedades dos filmes de BNT até 1.4 GHz, relativamente às propriedades medidas à frequência de 1 MHz. Após a prova de conceito, são conduzidos estudos de optimização do processo de deposição. Para tal foram especificamente sintetizados por processo convencional de reacções no estado sólido pós de BNT. O sucesso no processo de fabricação por EPD está intimamente relacionado com a escolha do meio de suspensão e aditivos, que devem originar uma suspensão estável com um grau de dispersão das partículas elevado. Neste trabalho são estudados quatro meios suspensores diferentes, que incluem, água, acetona, etanol e ácido acético. As propriedades físico – químicas das diferentes suspensões foram analisadas pelo determinação do potencial zeta, da distribuição do tamanho de partícula e transmitância de luz. Os resultados experimentais revelam que o potencial zeta é uma medida directa da estabilidade das suspensões, visto que o máximo do potencial zeta corresponde ao máximo de dispersão da suspensão que se reflecte também na distribuição do tamanho de partícula e no comportamento em termos de transmissão de luz através da suspensão. O máximo de potencial zeta (61 mV) é obtido para o meio suspensor de acetona com adições de I2, a que corresponde uma transmitância de 10%. O efeito dos diferentes meios suspensores é estudado na deposição, microestrutura e propriedades dieléctricas dos filmes de BNT. De entre os vários meios suspensores, apenas o ácido acético e a acetona com I2 apresentam a capacidade para formação de depósitos e as limitações do ácido acético são analisadas em termos da reproductibilidade do processo. Camadas depositadas de homogeneidade e taxa de deposição elevadas e com superfícies macias foram obtidas com o meio suspensor à base de acetona. Para este caso, o efeito de vários parâmetros de processamento, que incluem o campo eléctrico aplicado, o tempo de deposição e a composição da suspensão e a espessura e morfologia dos filmes é investigada e a discutida. Conjuntamente, e sob as condições optimizadas de deposição é estudado o efeito da temperatura de sinterização na estrutura, microestrutura e propriedades dieléctricas dos filmes espessos de BNT. Para tal filmes de BNT com 10 a 80 μm de espessura foram fabricados por EPD sobre folhas de Pt em diferentes condições. O impacto dos parâmetros de processamento: campo eléctrico, substrato e temperatura de sinterização são analisados e discutidos. Observa-se que um aumento da temperatura de sinterização aumenta acentuadamente a razão de aspecto dos grãos, decresce a permitividade dieléctrica relativa e o coeficiente de temperatura da permitividade TCεr varia de -114 para +12 ppm/°C. É então proposto que a anisotropia do grão observada é facilidade pelas condições de sinterização restrictas (“constrained sintering”). Através do controlo da temperatura de sinterização filmes espessos de TCεr quase nulo, Q elevado com 45< εr<70 podem ser fabricados. Esta descoberta é de especial relevância tecnológica visto que demonstra que o controlo da tensão originada pelo substrato e as condições de sinterização podem ser usadas para controlar a anisotropia do cerscimento do grão e consequentemente as propriedades dieléctricas dos filmes de BaO-Re2O3-TiO2. Ao mesmo tempo o fabrico de filmes com propriedades controladas contribui para a diminuição das dimensões dos dispositivos que operam a frequências elevadas. Esperam-se observações semelhantes em outros sistemas de materiais, o que abre ainda mais o leque de oportunidades em termos tecnológicos. Para aplicações específicas, camadas espessas de dieléctricos sobre substratos isoldaores podem ser necassárias. Contudo a fabricação de filmes por deposição electroforética é incompatível com o uso de substratos isoladores; assim sendo foi desenvolvido e é apresentado neste trabalho um método de fabricação de filmes sobre substratos isoladores por deposição electroforética. Para ultrapassar esta dificuldade, substratos de alumina foram recobertos com uma “camada sacríficio” de grafite. Filmes uniformes e densos foram então depositados sobre substratos de alumina (Al2O3). A influência da espessura da camada de grafite e a sua interacção com os filmes de BNT na estrutura, microestrutura e resposta dieléctrica é avaliada e discutida. Intercações marcadas entre os filems de BNT e os substratos de alumina foram observadas a temperaturas superiores a 1300 ºC. A difusão dos iões de Al para o interior dos filmes origina a formação de segundas fases de aluminatos de neodimio. Contudo filmes de BNT de espessura de 100 μm e sinterizados a 1250 ºC durante 1h exibem uma permitividade dieléctrica relativa e valores de Q de146 e 1161 a cerca de 10 GHz, respectivamente. Crê-se que esta aproximação do uso de uma “camada sacríficio” de grafite sobre substratos não condutores é muito importante e válida, já que pode ser extendida para a deposição de filmes espessos e finos a uma variedade alargada de materiais funcionais sobre um leque também alargado de substratos não condutores. Um outro desafio em termos de industria microelectónica, em particular a relacionada com o fabrico de dispositivos sintonizáveis de operação a frequências elevadas, é a fabricação de dieléctricos de perdas baixas e sintonabilidade elevada. Neste trabalho propõe-se uma aborgadem de engenharia para ultrapassar esta limitação. Assim reporta-se a preparação e caracterização de compósitos de BaNd2Ti5O14 (BNT) - Ba0.5Sr0.5TiO3 (BST) sobre folhas de Pt, através da combinação do processo de deposição electroforética com o processo sol gel. Filmes compósitos de BNT – BST com espessuras de 9μm sobre Pt, homogéneos, densos e uniformes exibem permitividade e perda dieléctrica de 287 e 0.0013 a 1MHz, e sintonização da permitividade dieléctrica de 12% a 33 kV/cm, e coeficiente de temperatura da permitividade de 0.26% entre 28 ºC e 120 ºC, respectivamente. Acima de tudo estes filmes exibem um dos facoters de qualidade K mais elevados referidos na literatura. Assim, a actual limitação de dieléctricos sintonizáveis de baixa perda é de certo modo ultrapassada e considera-se que estes resultados têm uma implicação alargada na comunidade microelectrónica de dispositivos sintonizáveis a frequências elevadas.
Currently RF and MW design engineers are being asked to send more bits over less bandwidth, use less battery power, and create products that are smaller and more flexible, which include increased integration, operation at higher frequencies and reduced costs. In this context the replacement of the current bulk ceramic dielectric components of some microelectronic devices (filters, baluns and antennas) by dielectrics processed as thick films is now being considered. With this methodology it is expected, besides reducing device size, to reduce processing costs and of particular relevancy are the opportunities created by the possibility to process thick films conformally on substrates and on metal foils. New structures and designs for such devices to operate at high frequencies are then expected. At the same time, this drives the search for fabrication processes for films materials to be mass-produced with repeatable performance at very low costs. The techniques for the preparation of thick films include tape casting, screen printing, jet printing or electrophoretic deposition (EPD). The importance of EPD comes from its unique features, such as the high flexibility and simplicity for application with various materials and combinations of materials, and on a wide range of shapes and 3D complex and porous structures, its costeffectiveness, and its ability to be scaled-up to the fabrication of large product volumes and sizes. In addition, when compared with the other methods, EPD enables the fabrication of highly uniform layers with an easy control of the layer thickness. As so, EPD matches well with the current considerations of the microelectronics industry of replacement of bulk dielectric components by dielectrics processed as thick films. Pertaining to materials, dielectrics which may be employed as microwave components must exhibit low loss or high quality factor Q, high relative permittivity and small temperature coefficient of relative permittivity (TCεr). Within low loss dielectrics, the BaO-Nd2O3-TiO2 system represents an important commercial family of microwave materials, particularly the 1:1:5 composition, that because it exhibits low dielectric losses, high quality factor, high relative permittivity and small temperature coefficient of resonant frequency, have been known as an important microwave dielectric material. Although BaO-Nd2O3-TiO2 ceramics are currently being produced for resonators, filters and substrates, among others applications, the use of BNT thick films have not been reported until the realization of this PhD program. In this work, the fabrication of BaNd2Ti5O14 (BNT) thick films by EPD is exploited. For that a systematic research study of the EPD process is conducted from the proof-of-concept till the definition of reproducible and optimised process conditions. Commercial and “home made” BaNd2Ti5O14 powders were used to fabricate BNT films on platinum foils and polycrystalline alumina substrates. The technological feasibility of using EPD for the fabrication of BNT thick films was firstly studied. 12 to 52 μm thick BNT films were fabricated from BNT commercial powders on platinum metallic foils by EPD. To improve the microstructure and density of the films a post deposition isostatic pressing step was used. The effect of film thickness on the dielectric properties at low frequencies was investigated. As the film thickness increases, the dielectric properties of BNT films approach those of BNT ceramics in terms of permittivity and loss tangent. 52 μm-thick BNT film sintered at 1300 °C for 1 h exhibit a dielectric constant and a loss tangent of 107 and 0.0006 (or Q of 1600) at 1 MHz, respectively. The variation in permittivity is less than 0.02 % at a bias voltage ±8 kV/cm. The change of film permittivity with the temperature within the range 30-120 °C is below +58.5 ppm/°C. Compared to at 1 MHz, the dielectric properties of 52 μm thickness BNT films do not show tremendously degradation till up to 1.4 GHz. After the proof-of-concept, the optimization of the EPD process was conducted. For that BNT powders were specifically synthesised by the conventional solid state reaction method. The success in the EPD process is intimately related to a careful choice of the suspension media and additives, which should lead to well-dispersed and stable suspensions for EPD. Four different suspension media, which included de-ionized water, acetone, ethanol, and glacial acetic acid (HAC), were studied. The physicochemical properties of the different suspensions were evaluated and analyzed by zeta potential, particle size distribution and light transmittance. Experimental results revealed that the zeta potential is a straightforward indication of the stability of these suspensions, since the maximum absolute zeta potential corresponds to a maximum of the suspension dispersibility, also reflected in the particle size distribution and suspension light transmittance behaviour. The maximum zeta potential was obtained for acetone with iodine suspensions (61 mV), and the corresponding transmittance was 10%. The effect of different solvents was studied on the deposition, microstructure and dielectric properties of BNT thick films. Among the used solvents, only the acetic acid and acetone with I2 based suspensions showed the ability of forming deposits and the limitations of acetic acid solvent was analyzed in terms of the process reproducibility. Deposits with homogeneous, smooth surface and high deposition yield were obtained upon adding I2 to the acetone based suspension. For the case of acetone with I2 suspensions, the effect of EPD process parameters such as deposition voltage, deposition time and suspension composition, deposited thickness of BNT and film morphology was investigated and discussed. In addition, under the definition of the optimal conditions for EPD BNT thick films, the effect of the post deposition sintering temperature was addressed on the structure, microstructure and dielectric properties of BNT thick films. For that 10 to 80 μm thick BaNd2Ti5O14 (BNT) films were fabricated by electrophor-etic deposition on Pt foils under different conditions. The impact of the processing parameters: electric field during EPD, the substrate effect and the sintering temperature were analyzed and discussed. It was observed that the increase of the sintering temperature increases markedly the aspect ratio of the grains, decreases the dielectric permittivity and TCεr changes from -114 to +12 ppm/°C. It was then proposed that the o bserved anisotropic grain growth is facilitated by the constrained sintering. By controlling the sintering temperature, near – zero TCεr, high Q thick films can be fabricated with 45< εr<70. These findings are of technological relevance since they demonstrate that control of substrate constraint and sintering conditions can be used to control grain anisotropy and thus microwave properties of the BaO-Re2O3-TiO2. The thick films facilitate scaling to small device sizes for high frequency operation. Similar observations are expected in other MW systems thus opening further technological opportunities. For some specific applications, such as multilayer microstrip for band-pass microwave filter applications, thick dielectric layers on insulating substrates may be required. However the fabrication by EPD is incompatible with the use of insulating substrates, so a method of performing EPD on non-conducting substrates was developed and is reported in this work. To overcome the requirement of a conducting substrate, insulating polycrystalline alumina substrates were covered with a sacrificial graphite layer. Uniform and dense BNT layers have been then deposited on alumina (Al2O3) substrate by EPD. The influence of the graphite layer thickness and the interactions between the BNT films and alumina substrates on the final structure, microstructure and dielectric response were addressed and discussed. Severe interactions between the BNT films and alumina substrates were observed for sintering temperatures >1300 °C. The diff usion of Al ions into the films resulted in the formation of neodymium aluminates second phases. However 100 μm thick BNT films sintered at 1250 ºC/1h show relative permittivity and Q values of 146 and 1161 at about 10 GHz, respectively. It is believed that this approach of using sacrificial graphite layers for EPD on non-conducting substrates is extremely valuable since it can be extended for both thin and thick film deposition on a large variety of other non-conducting substrates. Another challenge for the microelectronics agile / tunable industries is the fabrication of low loss tunable microwave dielectrics because lower loss tangents provide lower insertion loss in the device. In the work an approach to overcome this limitation is proposed. The preparation and characterization of BaNd2Ti5O14 (BNT) - Ba0.5Sr0.5TiO3 (BST) composite thick films on flexible platinum foil substrate, via an EPD process combined with a sol gel one was reported. Homogeneous, dense, and uniform 9μm-thick BNT-BST composite thick films on flexible Pt foils exhibit dielectric constants and loss tangent of 287 and 0.0013 at 1MHz, and dielectric tunability of 12% at 33kV/cm, and tempera-ture coefficient of relative permittivity of 0.26% between 28ºC to 120ºC, respectively. Above all these films exhibit one of the highest quality factor (K = 70) reported for dielectric films. As such the actual limitation of low loss high tunable dielectrics is somehow surmount and these results are expected to have broad implications in the community of microwave agile devices.

Magnetismus ist ein Phänomen, das eine wichtige Rolle in einer Vielfalt technischer Anwendungen spielt. Ohne den Einsatz magnetischer Effekte und Materialen wäre der heutzutage erreichte technische Fortschritt unmöglich, da viele grundlegende Techniken wie Stromerzeugung, elektrischer Antrieb, Informationsübertragung und viele andere auf magnetische bzw. elektromagnetische Phänomene zurückzuführen sind. Dabei haben die ferromagnetischen Materialen stets zur Effizienz von elektrischen und elektronischen Anwendungen beigetragen, weswegen an diesen Materialen auch entsprechend viel geforscht worden ist. Moderne Technologien, insb. Massenspeicher basieren oft auf Ferromagneten und erfordern daher die weitere Erforschung und Anpassung ihrer Eigenschaften. Für die Funktionalität von Hochgeschwindigkeitsgeräten spielt das dynamische Verhalten dünner magnetischer Schichten eine kritische Rolle. In dieser Arbeit wird die Magnetisierungsdynamik dünner Schichtelemente mittels zeitaufgelöster Weitfeld- Kerrmikroskopie untersucht. Dies ist ein aktuelles Thema, an dem in den letzten Jahren sehr intensiv gearbeitet wird. Allerdings sind viele für die Anwendungen sehr wichtige Details des magnetischen Schaltens wegen ihre Vielfältigkeit und Komplexität doch nicht vollständig untersucht und verstanden. In dieser Arbeit werden überwiegend experimentelle Ergebnisse vorgestellt, die einen zusätzlichen Beitrag zum aktuellen Wissenstand leisten. In einem ferromagnetischen Körper bilden sich Bereiche mit spontaner Magnetisierung, die man als Domänen bezeichnet. Die spontane Magnetisierung entsteht aufgrund der Spin-Spin Wechselwirkung, und die Domänen bilden sich aufgrund der Energieminimierung des magnetisierten Körpers. Langsame Magnetisierungsprozesse werden im Wesentlichen getragen von Domänenumordnungen und Domänengrenzenverschiebungen. Solche Prozesse bezeichnet man als quasistatisch, da sich der Körper durch deren Langsamkeit immer im Gleichgewicht oder zumindest sehr nahe daran befindet. Mit zunehmender Anregungsgeschwindigkeit gilt diese Annahme nicht mehr, da die Präzessionsbewegung der magnetischen Momente das Schaltverhalten in diesem Fall definiert. Die Untersuchung der Magnetisierungsdynamik setzt die Möglichkeit voraus, nicht-unterbrochene Prozesse beobachten zu können. Dieses Ziel kann mittels stroboskopischer Abbildung erreicht werden. Dabei wird derselbe Prozess kontinuierlich wiederholt (vorausgesetzt, dass die Prozesse sich reproduzierbar wiederholen lassen), und zu definierten Zeitpunkten werden die entsprechenden Kerraufnahmen gemacht. Dafür wird eine CCD Kamera mit einem Photoverstärker benutzt, welcher als optischer Schalter fungiert. Die Zeitauflösung dieses Systems und damit auch das Vermögen, die Hochfrequenzvorgänge abzubilden, beträgt 250 ps. Die Eigenschaften des magnetischen Umschaltens hängen stark von der Elementgeometrie ab. Diese Unterschiede sind auf unterschiedliche Entmagnetisierungsfaktoren, und damit auf Unterschiede in den effektiven Feldern zurückzuführen. Solche Unterschiede werden auf zwei Weisen initiiert: ein quadratisches Element wird entlang unterschiedlicher Richtungen (entlang der Seite und der Diagonalen) angeregt; die Form des Elementes wird zwischen Quadrat und Rechteck mit unterschiedlichen Seitenverhältnissen variiert. Die beobachteten Schaltvorgänge werden miteinander verglichen und die Ergebnisse dargestellt. Dabei werden auch die dynamischen Vorgänge immer mit den quasistatischen verglichen. Aus dem Vergleich folgt, dass ein steigendes Seitenverhältnis zur geringeren Schaltgeschwindigkeit führt, und dass die dabei entstehenden Domänen zunehmend komplexer werden. Dabei gibt es wesentliche Unterschiede zwischen den dynamischen und quasistatischen Domänen, vor allem in der Domänenwandstruktur. Das Schalten an sich unterscheidet sich auch sehr stark. Quasistatisches Schalten erfolgt überwiegend durch Domänenwandbewegung, während das dynamische Schalten durch inkohärente Rotation der Magnetisierung im ganzen Element erfolgt. Das Hochfrequenzverhalten am Prototypen eines Mikroinduktors wird untersucht. Der Induktor besteht aus vielen magnetischen Elementen, die eine induzierte uniaxiale Anisotropie besitzen. Diese ist bei der Hälfte der Elemente entlang des Magnetfeldes, und bei der anderen Hälfte senkrecht zum Magnetfeld der Spule ausgerichtet. Das dynamische Verhalten der beiden Elementtypen unterscheidet sich stark, vor allem die Ummagnetisierungsgeschwindigkeit. Diese Unterschiede können zu einer Phasenverschiebung im elektrischen Signal führen, was die Effizienz des Induktors senkt. Durch die Untersuchung der Magnetisierungsdynamik in Wechselfeldern unterschiedlicher Frequenz ist auch festgestellt worden, dass bis 100 MHz die Magnetisierungsvorgänge überwiegend durch Domänenwandbewegung erfolgen, während ab 200 MHz- Rotationsprozesse stattfinden.

A stacked microchannel heat sink was developed to provide efficient cooling for microelectronics devices at a relatively low pressure drop while maintaining chip temperature uniformity. Microfabrication techniques were employed to fabricate the stacked microchannel structure, and experiments were conducted to study its thermal performance. A total thermal resistance of less than 0.1 K/W was demonstrated for both counter flow and parallel flow configurations. The effects of flow direction and interlayer flow rate ratio were investigated. It was found that for the low flow rate range the parallel flow arrangement results in a better overall thermal performance than the counter flow arrangement; whereas, for the large flow rate range, the total thermal resistances for both the counter flow and parallel flow configurations are indistinguishable. On the other hand, the counter flow arrangement provides better temperature uniformity for the entire flow rate range tested. The effects of localized heating on the overall thermal performance were examined by selectively applying electrical power to the heaters. Numerical simulations were conducted to study the conjugate heat transfer inside the stacked microchannels. Negative heat flux conditions were found near the outlets of the microchannels for the counter flow arrangement. This is particularly evident for small flow rates. The numerical results clearly explain why the total thermal resistance for counter flow arrangement is larger than that for the parallel flow at low flow rates. In addition, laminar flow inside the microchannels were characterized using Micro-PIV techniques. Microchannels of different width were fabricated in silicon, the smallest channel measuring 34 mm in width. Measurements were conducted at various channel depths. Measured velocity profiles at these depths were found to be in reasonable agreement with laminar flow theory. Micro-PIV measurement found that the maximum velocity is shifted significantly towards the top of the microchannels due to the sidewall slope, a common issue faced with DRIE etching. Numerical simulations were conducted to investigate the effects of the sidewall slope on the flow and heat transfer. The results show that the effects of large sidewall slope on heat transfer are significant; whereas, the effects on pressure drop are not as pronounced.

For microelectronics used in the low-temperature applications, the understanding of their reliability and performance has become an important research subject characterised as electronics to serve under the severe or extreme service conditions. Along with the impact from the increased miniaturization of devices, the various properties and the relevant thermo-mechanical response of the interconnection materials to temperature excursion at micro-scale become a critical factor which can affect the reliable performance of microelectronics in various applications. Pure indium as an excellent interconnection material has been used in pixellated detector systems, which are required to be functional at cryogenic temperatures. This thesis presents an extensive investigation into the thermo-mechanical properties of indium joints as a function of microstructure, strain (loading histories-dependent) and temperature (service condition-sensitive), specifically in the areas as follows: (i) the interfacial reactions and evolution between indium and substrate during the reflow process (liquid-solid) and thermal aging (solid-solid) stages by taking low-temperature cycling into account; (ii) determination of the effects of joint thickness and the types of substrate (e.g. Cu or Ni) on the mechanical properties of indium joints, and the stress- and temperature-dependent creep behaviour of indium joints; (iii) the establishment of a constitutive relationship for indium interconnects under a wide range of homologous temperature changes that was subsequently implemented into an FE model to allow the analysis of the evolution of thermally-induced stresses and strains associated with a hybrid pixel detector.

In microelectronics, the device size continues to shrink to improve the performance and functionality, which sets technical challenges for the integrated circuit (IC) fabrication. Novel materials and processing techniques are developed to maintain excellent device performances and structural reliability. Cobalt-based thin films possess numerous applications in microelectronics with the potential to enhance the device performance and reliability. This thesis explores the fabrication, characterization and application of cobalt-based thin films for microelectronics. Chemical vapor deposition (CVD) technique has been applied for depositing cobalt-based thin films, because CVD can produce high quality thin films with excellent conformality in complex 3D architectures required for future microelectronics.
Engineering and Applied Sciences

This thesis explores mechanical behavior of microelectronic devices and lithium-ion batteries. We first examine electromigration-induced void formation in solder bumps by constructing a theory that couples electromigration and creep. The theory can predict the critical current density below which voids do not form. Due to the effects of creep, this quantity is found to be independent of the solder size and decrease exponentially with increasing temperature, different from existing theories.
Engineering and Applied Sciences