Effects of the epitaxial layer thickness on the noise properties of Schottky barrier diodes

An analytical theory to describe the combined effects of the epitaxial layer thickness and the ohmic contact on the noise properties of Schottky barrier diodes is presented. The theory, which provides information on both the local and the global noise properties, takes into account the finite size of the epitaxial layer and the effects of the back ohmic contact, and applies to the whole range of applied bias. It is shown that by scaling down the epitaxial layer thickness, the current regime in which the noise temperature displays a shot-noise-like behavior increases at the cost of reducing the current range in which the thermal-noise-like behavior dominates. This improvement in noise temperature is limited by the effects of the ohmic contact, which appear for large currents. The theory is formulated on general trends, allowing its application to the noise analysis of other semiconductor devices operating under strongly inhomogeneous distributions of the electric field and charge concentrations.

Suitability of foam coating on application of thin liquid films

This work aimed to find out the suitability of foam as medium in application of thin liquid films. This consists of research over phenomena related to foam physics and behaviour. Solutions and mixtures to be foamed, foaming agents, foam generation and application methods were evaluated. Over the evaluated solutions and mixtures coating paste and CMC did not foam well. Latex and PVA solutions were foamable and the best solution for foam use was starch. PVA and casein can be used as foaming agents, but the best results were achieved with sodium dodecyl sulphate (SDS). SDS works well with starch solutions producing fine and stable foam. Foaming was done with simple mixers where pressurized air was fed to the solution. The foaming works fine when enough shear force is used together with sufficient foaming agent concentration. Foam application with curtain, rod and cylinder methods with a gap between the application device and paper were not usable because of high coating amount. Coating amounts were smallest with the blade method which achieved 0.9 g/m2 starch layer. Although some strength decrease was expected because of the foaming agent, it dit not have significant effect. The targeted coating amount of 0.5 g/m2 was not achieved due to the limitations with the methods. More precise foam application methods are needed. Continuous foam generation and feed to the paper surface with controllable device such as application teeth could improve the results.

Polymer surface modification by atomic layer deposition

Current industrial atomic layer deposition (ALD) processes are almost wholly confined to glass or silicon substrates. For many industrial applications, deposition on polymer substrates will be necessary. Current deposition processes are also typically carried out at temperatures which are too high for polymers. If deposition temperatures in ALD can be reduced to the level applicable for polymers, it will open new interesting areas and applications for polymeric materials. The properties of polymers can be improved for example by coatings with functional and protective properties. Although the ALD has shown its capability to operate at low temperatures suitable for polymer substrates, there are other issues related to process efficiency and characteristics of different polymers where new knowledge will assist in developing industrially conceivable ALD processes. Lower deposition temperature in ALD generally means longer process times to facilitate the self limiting film growth mode characteristic to ALD. To improve process efficiency more reactive precursors are introduced into the process. For example in ALD oxide processes these can be more reactive oxidizers, such as ozone and oxygen radicals, to substitute the more conventionally used water. Although replacing water in the low temperature ALD with ozone or plasma generated oxygen radicals will enable the process times to be shortened, they may have unwanted effects both on the film growth and structure, and in some cases can form detrimental process conditions for the polymer substrate. Plasma assistance is a very promising approach to improve the process efficiency. The actual design and placement of the plasma source will have an effect on film growth characteristics and film structure that may retard the process efficiency development. Due to the fact that the lifetime of the radicals is limited, it requires the placement of the plasma source near to the film growth region. Conversely this subjects the substrate to exposure byother plasma species and electromagnetic radiation which sets requirements for plasma conditions optimization. In this thesis ALD has been used to modify, activate and functionalize the polymer surfaces for further improvement of polymer performance subject to application. The issues in ALD on polymers, both in thermal and plasma-assisted ALD will be further discussed.

MODELING A COMPUTER APPLICATION FOR MANAGEMENT OF MAINTENANCE ACTIVITIES OF UNPAVED ROADS

ABSTRACTThis study presents a contribution to the modeling of a computer application employing a method of serviceability performance for unpaved roads, aiming the management of maintenance/restoration activities of the primary surface layer. The proposed methodology consisted of field inspections during dry (April to September) and rainy (October to March) periods, during which objective evaluations were performed to survey of defects and their densities and degrees of severity. To aid the functional classification of analyzed road sections and the determination of the defect with major influence on the serviceability of these roads, the method of serviceability performance proposed by Silva (2009)was implemented in the Visual Basic for Applications (VBA) language in Microsoft Excel software. With the use of the computer application proposed it was possible to identify among the defects analyzed in field, through the index of serviceability of the sampling unit per defect type (ISUdef), which one had the greatest influence on determining the relative serviceability index per road section (IST). The results allow us to conclude that the computer application Road achieved satisfactory results, since the objective evaluation criteria applied to road sections denotes consistency regarding their serviceability.

Recycling of nutrients with application of organic waste in degraded pasture

The utilization of organic wastes represents an alternative to recover degraded pasture. The experiment aimed to assess the changes caused by the provision of different organic waste (poultry litter, turkey litter and pig manure) in a medium-textured Oxisol in Brazilian Savanna under degraded pasture. It was applied different doses of waste compared to the use of mineral fertilizers and organic mineral and evaluated the effect on soil parameters (pH, organic matter, phosphorus and potassium) and leaf of Brachiariadecumbens (crude protein, phosphorus and dry mass production). It was observed that application of organic waste did not increase the level of soil organic matter and pH in the surface layer, and the application of turkey litter caused acidification at depths of 0.20-0.40 m and 0.40-0.60 m. There was an increase in P and K in the soil with the application of poultry litter and swine manure. All organic wastes increased the productivity of dry matter and crude protein and phosphorus. The recycling of nutrients via the application of organic waste allows efficiency of most parameters similar to those observed with the use of mineral sources, contributing to improving the nutritional status of soil-plantsystem.

Effects of vinasse application under the physical attributes of soil covered with sugarcane straw

Areas under vinasse application have been associated to favorable physical conditions for root development, aeration, infiltration and water movement in soil profile. This study aimed to evaluate changes on physical attributes of soil under sugarcane straw after vinasse application in two sugarcane growing areas (Area 1 and Area 2) under mechanized management in the state of Paraíba, Brazil. In each area, the samples were collected in the 0-0.20, 0.20-0.40 and 0.40-0.60m layers of the soil, in 36 points, distributed in a 10×10m mesh, one day before and 40 days after vinasse application. The data were submitted to multivariate analysis with repeated measures and geostatistics. The vinasse application decreased soil density and increased total porosity in both Areas and increased organic matter in Area 2. In Area 1 occurred pure nugget effect for the fractions of sand, silt and clay, independent of soil layer. In Area 2, this effect was verified mostly at superficial layers, except for the fraction of clay that presented a moderate degree of spatial dependence.

Application of x-ray computed tomography in the evaluation of soil porosity in soil management systems

The study aimed to evaluate a methodology to quantify the porosity of the soil using computed tomography in areas under no-tillage, conventional tillage and native forest. Three soil management systems were selected for the study: forest, conventional tillage and no-tillage. In each soil management system, undisturbed soil samples were collected in the surface layer (0.0 to 0.10 m). The tomographic images were obtained using a X-ray microtomography. After obtaining the images, they were processed, and a methodology was evaluated for image conversion into numerical values. The statistical method which provided the greatest accuracy was the percentile method. The methodology used to analyze the tomographic image allowed quantifying the porosity of the soil under different soil management. The method enabled the characterization of soil porosity in a non-evasive and non-destructive way.

Non-Linear Unsteady Aerodynamic Response Approximation Using Multi-Layer Functionals

Non-linear functional representation of the aerodynamic response provides a convenient mathematical model for motion-induced unsteady transonic aerodynamic loads response, that accounts for both complex non-linearities and time-history effects. A recent development, based on functional approximation theory, has established a novel functional form; namely, the multi-layer functional. For a large class of non-linear dynamic systems, such multi-layer functional representations can be realised via finite impulse response (FIR) neural networks. Identification of an appropriate FIR neural network model is facilitated by means of a supervised training process in which a limited sample of system input-output data sets is presented to the temporal neural network. The present work describes a procedure for the systematic identification of parameterised neural network models of motion-induced unsteady transonic aerodynamic loads response. The training process is based on a conventional genetic algorithm to optimise the network architecture, combined with a simplified random search algorithm to update weight and bias values. Application of the scheme to representative transonic aerodynamic loads response data for a bidimensional airfoil executing finite-amplitude motion in transonic flow is used to demonstrate the feasibility of the approach. The approach is shown to furnish a satisfactory generalisation property to different motion histories over a range of Mach numbers in the transonic regime.

Development of 3D super-resolution tomographic STED microscopy and its application to studies on bone resorption

In this doctoral thesis, a tomographic STED microscopy technique for 3D super-resolution imaging was developed and utilized to observebone remodeling processes. To improve upon existing methods, wehave used a tomographic approach using a commercially available stimulated emission depletion (STED) microscope. A certain region of interest (ROI) was observed at two oblique angles: one at a standard inverted configuration from below (bottom view) and another from the side (side view) via a micro-mirror positioned close to the ROI. The two viewing angles were reconstructed into a final tomogram. The technique, named as tomographic STED microscopy, was able to achieve an axial resolution of approximately 70 nm on microtubule structures in a fixed biological specimen. High resolution imaging of osteoclasts (OCs) that are actively resorbing bone was achieved by creating an optically transparent coating on a microscope coverglass that imitates a fractured bone surface. 2D super-resolution STED microscopy on the bone layer showed approximately 60 nm of lateral resolution on a resorption associated organelle allowing these structures to be imaged with super-resolution microscopy for the first time. The developed tomographic STED microscopy technique was further applied to study resorption mechanisms of OCs cultured on the bone coating. The technique revealed actin cytoskeleton with specific structures, comet-tails, some of which were facing upwards and some others were facing downwards. This, in our opinion, indicated that during bone resorption, an involvement of the actin cytoskeleton in vesicular exocytosis and endocytosis is present. The application of tomographic STED microscopy in bone biology demonstrated that 3D super-resolution techniques can provide new insights into biological 3D nano-structures that are beyond the diffraction-limit when the optical constraints of super-resolution imaging are carefully taken into account.

Hydrogen Sensor Application of Anodic Titanium Oxide Nanostructures

Hydrogen (H2) fuel cells have been considered a promising renewable energy source. The recent growth of H2 economy has required highly sensitive, micro-sized and cost-effective H2 sensor for monitoring concentrations and alerting to leakages due to the flammability and explosiveness of H2 Titanium dioxide (TiO2) made by electrochemical anodic oxidation has shown great potential as a H2 sensing material. The aim of this thesis is to develop highly sensitive H2 sensor using anodized TiO2. The sensor enables mass production and integration with microelectronics by preparing the oxide layer on suitable substrate. Morphology, elemental composition, crystal phase, electrical properties and H2 sensing properties of TiO2 nanostructures prepared on Ti foil, Si and SiO2/Si substrates were characterized. Initially, vertically oriented TiO2 nanotubes as the sensing material were obtained by anodizing Ti foil. The morphological properties of tubes could be tailored by varying the applied voltages of the anodization. The transparent oxide layer creates an interference color phenomena with white light illumination on the oxide surface. This coloration effect can be used to predict the morphological properties of the TiO2 nanostructures. The crystal phase transition from amorphous to anatase or rutile, or the mixture of anatase and rutile was observed with varying heat treatment temperatures. However, the H2 sensing properties of TiO2 nanotubes at room temperature were insufficient. H2 sensors using TiO2 nanostructures formed on Si and SiO2/Si substrates were demonstrated. In both cases, a Ti layer deposited on the substrates by a DC magnetron sputtering method was successfully anodized. A mesoporous TiO2 layer obtained on Si by anodization in an aqueous electrolyte at 5°C showed diode behavior, which was influenced by the work function difference of Pt metal electrodes and the oxide layer. The sensor enabled the detection of H2 (20-1000 ppm) at low operating temperatures (50–140°C) in ambient air. A Pd decorated tubular TiO2 layer was prepared on metal electrodes patterned SiO2/Si wafer by anodization in an organic electrolyte at 5°C. The sensor showed significantly enhanced H2 sensing properties, and detected hydrogen in the range of a few ppm with fast response/recovery time. The metal electrodes placed under the oxide layer also enhanced the mechanical tolerance of the sensor. The concept of TiO2 nanostructures on alternative substrates could be a prospect for microelectronic applications and mass production of gas sensors. The gas sensor properties can be further improved by modifying material morphologies and decorating it with catalytic materials.

Deep learning of representations and its application to computer vision

L’objectif de cette thèse par articles est de présenter modestement quelques étapes du parcours qui mènera (on espère) à une solution générale du problème de l’intelligence artificielle. Cette thèse contient quatre articles qui présentent chacun une différente nouvelle méthode d’inférence perceptive en utilisant l’apprentissage machine et, plus particulièrement, les réseaux neuronaux profonds. Chacun de ces documents met en évidence l’utilité de sa méthode proposée dans le cadre d’une tâche de vision par ordinateur. Ces méthodes sont applicables dans un contexte plus général, et dans certains cas elles on tété appliquées ailleurs, mais ceci ne sera pas abordé dans le contexte de cette de thèse. Dans le premier article, nous présentons deux nouveaux algorithmes d’inférence variationelle pour le modèle génératif d’images appelé codage parcimonieux “spike- and-slab” (CPSS). Ces méthodes d’inférence plus rapides nous permettent d’utiliser des modèles CPSS de tailles beaucoup plus grandes qu’auparavant. Nous démontrons qu’elles sont meilleures pour extraire des détecteur de caractéristiques quand très peu d’exemples étiquetés sont disponibles pour l’entraînement. Partant d’un modèle CPSS, nous construisons ensuite une architecture profonde, la machine de Boltzmann profonde partiellement dirigée (MBP-PD). Ce modèle a été conçu de manière à simplifier d’entraînement des machines de Boltzmann profondes qui nécessitent normalement une phase de pré-entraînement glouton pour chaque couche. Ce problème est réglé dans une certaine mesure, mais le coût d’inférence dans le nouveau modèle est relativement trop élevé pour permettre de l’utiliser de manière pratique. Dans le deuxième article, nous revenons au problème d’entraînement joint de machines de Boltzmann profondes. Cette fois, au lieu de changer de famille de modèles, nous introduisons un nouveau critère d’entraînement qui donne naissance aux machines de Boltzmann profondes à multiples prédictions (MBP-MP). Les MBP-MP sont entraînables en une seule étape et ont un meilleur taux de succès en classification que les MBP classiques. Elles s’entraînent aussi avec des méthodes variationelles standard au lieu de nécessiter un classificateur discriminant pour obtenir un bon taux de succès en classification. Par contre, un des inconvénients de tels modèles est leur incapacité de générer deséchantillons, mais ceci n’est pas trop grave puisque la performance de classification des machines de Boltzmann profondes n’est plus une priorité étant donné les dernières avancées en apprentissage supervisé. Malgré cela, les MBP-MP demeurent intéressantes parce qu’elles sont capable d’accomplir certaines tâches que des modèles purement supervisés ne peuvent pas faire, telles que celle de classifier des données incomplètes ou encore celle de combler intelligemment l’information manquante dans ces données incomplètes. Le travail présenté dans cette thèse s’est déroulé au milieu d’une période de transformations importantes du domaine de l’apprentissage à réseaux neuronaux profonds qui a été déclenchée par la découverte de l’algorithme de “dropout” par Geoffrey Hinton. Dropout rend possible un entraînement purement supervisé d’architectures de propagation unidirectionnel sans être exposé au danger de sur- entraînement. Le troisième article présenté dans cette thèse introduit une nouvelle fonction d’activation spécialement con ̧cue pour aller avec l’algorithme de Dropout. Cette fonction d’activation, appelée maxout, permet l’utilisation de aggrégation multi-canal dans un contexte d’apprentissage purement supervisé. Nous démontrons comment plusieurs tâches de reconnaissance d’objets sont mieux accomplies par l’utilisation de maxout. Pour terminer, sont présentons un vrai cas d’utilisation dans l’industrie pour la transcription d’adresses de maisons à plusieurs chiffres. En combinant maxout avec une nouvelle sorte de couche de sortie pour des réseaux neuronaux de convolution, nous démontrons qu’il est possible d’atteindre un taux de succès comparable à celui des humains sur un ensemble de données coriace constitué de photos prises par les voitures de Google. Ce système a été déployé avec succès chez Google pour lire environ cent million d’adresses de maisons.

Optimization of Two Stage Process for the Growth of In2S3 , CuInSe2 and CuIn (Sel-xSx)2Thin Films for Solar Cell Application

Two stage processes consisting of precursor preparation by thermal evaporation followed by chalcogenisation in the required atmosphere is found to be a feasible technique for the PV materials such as n-Beta In2S3, p-CulnSe2, p-CulnS2 and p-CuIn(Sel_xSx)2. The growth parameters such as chalcogenisation temperature and duration of chalcogenisation etc have been optimised in the present study.Single phase Beta-In2S3 thin films can be obtained by sulfurising the indium films above 300°C for 45 minutes. Low sulfurisation temperatures required prolonged annealing after the sulfurisation to obtain single phase Beta-1n2S3, which resulted in high material loss. The maximum band gap of 2.58 eV was obtained for the nearly stoichiometric Beta-In2S3 film which was sulfurised at 350°C. This wider band gap, n type Beta-In2S3 can be used as an alternative to toxic CdS as window layer in photovoltaics .The systematic study on the structural optical and electrical properties of CuInSe2 films by varying the process parameters such as the duration of selenization and the selenization temperature led to the conclusion that for the growth of single-phase CuInSe2, the optimum selenization temperature is 350°C and duration is 3 hours. The presence of some binary phases in films for shorter selenization period and lower selenization temperature may be due to the incomplete reaction and indium loss. Optical band gap energy of 1.05 eV obtained for the films under the optimum condition.In order to obtain a closer match to the solar spectrum it is desirable to increase the band gap of the CulnSe2 by a few meV . Further research works were carried out to produce graded band gap CuIn(Se,S)2 absorber films by incorporation of sulfur into CuInSe2. It was observed that when the CulnSe2 prepared by two stage process were post annealed in sulfur atmosphere, the sulfur may be occupying the interstitial positions or forming a CuInS2 phase along with CuInSe2 phase. The sulfur treatment during the selenization process OfCu11 ln9 precursors resulted in Culn (Se,S)2 thin films. A band gap of 1.38 eV was obtained for the CuIn(Se,S)2.The optimised thin films n-beta 1n2S3, p-CulnSe2 and p-Culn(Sel-xSx)2 can be used for fabrication of polycrystalline solar cells.