Tesi sul tema "Electrochemical silicon etching"

In this project, nanopore arrays have been fabricated on bulk silicon and on silicon membranes by electrochemical etching. First, the surface of bulk silicon and silicon membranes have been patterned by photolithography and then invert pyramidal pit arrays have been formed by KOH etching. To fabricate nanopore arrays, bulk silicon and silicon membranes with the inverted pyramidal structure were electrochemically etched with backside illumination and by breakdown methods, respectively. Pore morphology was then characterized by scanning electron microscopy (SEM). On bulk silicon, etching by backside illumination did not form promising nanopore arrays; while arrays of nanopores with ~8 nm in diameter have been fabricated to a depth of 18 μm by tuning the applied breakdown bias. On silicon membranes, arrays of nanopores with 18±4 nm diameter have been etched through the membranes with the buried oxide remaining on the backside using the breakdown method.

Le Silicium est le deuxième élément le plus abondant dans la croûte terrestre après l’oxygène. Il est produit par voie métallurgique dans un four à arc électrique, le quartz est réduit en présence de réducteurs (charbon de bois, houille et coke de pétrole). Le silicium métallurgique est principalement utilisé dans la métallurgie comme élément d’alliage, dans la chimie et l’industrie solaire. Le prix du Silicium est fonction de sa pureté. Les travaux de cette thèse se divisent en deux parties l’utilisation du Silicium Métallurgique (99% Si) pour le stockage de l’hydrogène, et la photoluminescence du ferrosilicium (disiliciure de fer) de qualité métallurgique. Des substrats de silicium métallurgique ont été soumis à une anodisation électrochimique dans une solution à base d’acide fluorhydrique. Le silicium poreux nanostructuré obtenu est légèrement différent du silicium poreux issu de substrat de silicium de qualité électronique de même résistivité. L’influence des principaux paramètres sur la génération de l’hydrogène : la porosité, la concentration, le volume et la température ont fait l’objet d’une étude détaillée. Le silicium poreux produit à partir de silicium métallurgique est un matériau de stockage d’hydrogène. Des substrats de disiliciure de fer de qualité métallurgique ont été soumis à une anodisation électrochimique. Le composé obtenu est du disiliciure de fer nanostructuré avec du silicium résiduel, ce produit est recouvert de fluorosilicate de fer hexahydraté qui a la particularité d’être luminescent. Il s’agit à ce jour de la première anodisation du disiliciure de fer, un mécanisme de gravure a été proposé et l’influence des principaux paramètres d’anodisation sur les propriétés de photoluminescence a été évaluée
Silicon is the second most abundant element in the Earth crust after oxygen. Its use in metallurgy, building and electronic industry requires a huge fabrication level. Depending on the contamination level allowed, the price of this material varies in the orders of magnitude. This thesis focuses on the use of dirtiest metallurgical grade silicon and iron disilicide substrates for hydrogen storage and photoluminescence applications. The initial substrates were subjected to electrochemical etching in hydrofluoric acid-containing solutions. Anodization of metallurgical grade silicon substrate produces nanostructured porous silicon with somewhat shifted parameters (comparing with electronic grade porous silicon with the same resistivity), as it was studied in this thesis in details. It was shown, that metallurgical grade porous silicon can be applied as hydrogen storage material. Hydrogen generation is studied here based on the influences of some technically critical parameters: porosity, alkali concentration, volume and temperature. Electrochemical treatment of metallurgical grade iron disilicide substrates produces luminescent iron fluorosilicate hexahydrate, covering the residual nanostructured iron disilicide/silicon. Here, the influence of anodization parameters on photoluminescent properties is studied. Also, etching mechanism is proposed as for the new material never anodized

Los cristales fotónicos son materiales creados artificialmente, que pueden hacer con los fotones lo que los semiconductores ordinarios hacen con los electrones: es decir, pueden mostrar una banda fotónica prohibida (PBG), situación en la cual fotones con determinadas energías no pueden propagarse dentro del cristal independientemente de la polarización y la dirección de propagación. Por lo tanto, la banda prohibida para los fotones puede ser el verdadero análogo óptico de la banda prohibida fundamental en los semiconductores. Desde su invento en 1987, los cristales fotónicos han atraído un interés considerable debido a sus propiedades ópticas inusuales. Las propiedades únicas de los cristales fotónicos también han llevado al reconocimiento de su estudio como un nuevo y principal campo de la optoelectrónica.
El silicio macroporoso con su elevada constante dieléctrica, sus altas relaciones de aspecto y su total compatibilidad con la industria microelectrónica es un modelo excelente para estudiar las propiedades ópticas de cristales fotónicos bidimensionales y asimismo tridimensionales. Adicionalmente, se ha demostrado que el silicio macroporoso tiene varias aplicaciones únicas en muchos otros campos, como la electrónica, el micromecanizado, la detección de gases y la biotecnología. La investigación del silicio macroporoso crece continuamente debido a su enorme potencial de aplicaciones.
El trabajo presentado en esta tesis trata dos temas: simulación de la estructura de bandas fotónicas y análisis de cristales fotónicos bidimensionales, y la fabricación de estructuras bidimensionales basadas en silicio macroporoso para aplicaciones como cristales fotónicos en el espectro infrarrojo.
Debido a que muchas posibles aplicaciones de los cristales fotónicos están basadas
en sus bandas fotónicas prohibidas, es interesante diseñar cristales fotónicos con una banda
prohibida absoluta, que sea tan grande como es posible. En esta tesis describimos el método para alargar la banda fotónica absoluta, mostrando el papel de la simetría en el diseño de estructuras fotónicas óptimas. Hemos estudiado como reduciendo la simetría mediante incorporación de elementos adicionales en la celda unitaria o mediante cambio de la forma de los "átomos" afecta la relación de dispersión de los dos modos de polarización (TM y TE) en cristales fotónicos bidimensionales. Nuestro objetivo ha sido optimizar la magnitud de la banda fotónica absoluta, reduciendo la simetría de las celdas cuadrada y triangular y construir de este modo estructuras nuevas, llamadas celdas híbridas. Usando el método de las deferencias finitas en el dominio de tiempo (FDTD) hemos realizado un detallado análisis numérico de la relación de dispersión en celdas híbridas bidimensionales que consisten en columnas de aire en silicio.
En el caso de celda cuadrada, la reducción de la simetría ha sido aplicada con éxito para maximizar la magnitud de la banda prohibida. Para la celda cuadrada que consiste en columnas cilíndricas de aire, la incorporación de una columna adicional aumenta tres veces la magnitud de la PBG absoluta. En el caso de celda cuadrada de columnas cuadradas de aire, la rotación de las columnas juega un papel crítico en la creación de la PBG absoluta.
Si las columnas cuadradas no están rotadas no existe una PBG absoluta. La magnitud de la PBG absoluta se ha mejorado considerablemente a través de la combinación de incorporación de una columna adicional y rotación de las columnas cuadradas. Además, se genera una nueva PBG absoluta que se encuentra para un amplio rango de ángulos de rotación y dimensiones de las columnas, que están lejos de la condición de empaquetamiento (cuando las columnas se tocan). Esto favorece la fabricación de los cristales fotónicos.
La PBG absoluta es de mayor magnitud para la celda triangular formada por columnas cilíndricas de aire. Los resultados de las simulaciones demuestran que modificando la estructura triangular mediante incorporación de columnas adicionales o mediante columnas cuadradas (aunque las columnas estén rotadas) no mejora la PBG absoluta, por lo menos en el caso estudiado de estructura aire/silicio. La adición de columnas adicionales en la celda triangular reduce la magnitud de la PBG absoluta.
Hemos realizado un detallado análisis cuantitativo de las PBG absolutas para 2D celdas triangulares y hexagonales, considerando que entre las columnas y la matriz dieléctrica hay una capa superficial de otro material dieléctrico. Esta capa superficial puede ser indeseada (resultado del proceso de fabricación) o puede ser creada intencionadamente.
Las propiedades de las bandas fotónicas se ven afectadas del grosor y también de la constante dieléctrica de la capa superficial. Los resultados de las simulaciones demuestran que para estructuras que están formadas por columnas de aire en un material dieléctrico la existencia de una capa superficial reduce la magnitud de la PBG absoluta. Por otro lado, para estructuras formadas de columnas dieléctricas en aire la capa superficial puede mejorar la PBG cuando la constante dieléctrica de la capa es mayor que la de las columnas.
Esto proporciona mayor flexibilidad en la realización práctica de estos 2D cristales fotónicos. Por ejemplo, en ciertas ocasiones es imposible obtener pilares dieléctricos con un diámetro determinado o de un material concreto por limitaciones tecnológicas. Sin embargo, los pilares se pueden fabricar de un material con menor constante dieléctrica para el cual existe una técnica bien desarrollada. Después los pilares se pueden cubrir con el material deseado mediante deposición, obteniendo las mismas propiedades como en el caso de la estructura sin capa superficial.
Hemos desarrollado un equipo de ataque electroquímico para fabricación de 2D estructuras periódicas basadas en la formación de silicio macroporoso. Asimismo, hemos realizado un estudio de la influencia de los parámetros del ataque electroquímico sobre la morfología de los poros. Crecimiento estable de macroporos se puede obtener sólo si todos los parámetros del proceso de ataque (resistividad del substrato, concentración de HF, corriente de ataque, potencial anódico, temperatura, etc.) están ajustados apropiadamente.
Las condiciones óptimas ocupan una pequeña parte de todos los posibles parámetros del proceso. Por ejemplo, concentraciones de HF mayores de 10 wt.%, que se usan generalmente para crecer películas micro- y mesoporosas, no son apropiadas para crecer macroporos con una profundidad grande y una forma cilíndrica. Potenciales relativamente altos (para nuestras muestras mayores de 2 V) aumentan inevitablemente la formación de "breakdown-type" poros. Por otro lado, potenciales relativamente bajos (menores de 1 V) generalmente producen un crecimiento inestable de los poros que están parcial o totalmente recubiertos de silicio microporoso.
La corriente aplicada es el parámetro más crítico del proceso. Densidades de la corriente mayors de la densidad crítica Jps, que depende de la temperatura y de la concentración de HF, situaría el proceso en la región de electropulido. El control de la corriente durante el proceso es una tarea clave. Mantener la corriente de ataque constante durante todo el proceso es insuficiente para el crecimiento estable de macroporos cilíndricos. Se han identificado dos efectos que influyen la forma de los poros en profundidad. Primero, la concentración de HF disminuye cerca de la punta de los poros debido a las limitaciones por difusión en poros estrechos y hondos. Este efecto produce un incremento del diámetro del poro cerca de la punta. Segundo, la superficie interna de los poros aumenta para prolongados tiempos de ataque, provocando un incremento de la corriente de oscuridad y por lo tanto la formación de poros cónicos. Su diámetro decrece en profundidad. El incremento de la corriente de ataque de manera adecuada, tal que se produzca crecimiento de poros con forma cónica inversa, es un método para compensar la conicidad inicial de los poros. Si el ataque se realiza a temperaturas más bajas y burbujeando el electrolito con nitrógeno se puede reducir la corriente de oscuridad, formando poros menos cónicos. Otro método efectivo es el uso de un surfactante apropiado. Los surfactantes se usan por lo general para prevenir degradación causada por las burbujas de hidrógeno que se pegan en la superficie de la muestra. Hemos probado dos diferentes tipos de surfactants (TritonX-100 no iónico y SDS aniónico). Hemos observado que la adición de surfactantes no iónicos aumenta la corriente de oscuridad y la formación de poros cónicos. Por otro lado, el uso de surfactantes aniónicos reduce considerablemente la corriente de oscuridad y poros cilíndricos se pueden producir casi sin dificultad.
Aplicando las reglas explicadas arriba se han obtenido matrices altamente uniformes de macroporos con diferente distribución y dimensiones.
Por último, también se presentan algunos resultados preliminares sobre aplicaciones novedosas de silicio macroporoso. Las características estructurales de las matrices de macroporos se han utilizado para fabricar pilares de óxido de silicio que podrían encontrar aplicaciones en la biotecnología como plataformas tridimensionales para detección de reconocimiento de moléculas o como matrices de microjeringas. También se ha fabricado un filtro que consiste en membranas de silicio macroporoso y se han medido sus características ópticas. Este filtro se comporta como pasabajas cuando la luz incidente es paralela a los poros. Los resultados obtenidos son solamente cuantitativos y sugieren una futura optimización del proceso de ataque para fabricar muestras de alta calidad.
Asimismo se ha introducido modulación periódica del diámetro de los poros en profundidad y se han fabricado matrices de "ratchet-type" macroporos, los cuales podrían tener aplicaciones como dispositivos para separación de partículas. Se ha demostrado que mediante unos pocos pasos adicionales las matrices de macroporos modulados se pueden convertir en microestructuras tridimensionales de huecos interconectados. Esta técnica se puede aplicar para la fabricación de cristales fotónicos tridimensionales.
Photonic crystals are artificially created materials that can do to photons what an ordinary semiconductor does to electrons: that is to say, they can exhibit a photonic band gap, a situation in which photons with certain energies cannot propagate inside the crystal, regardless of polarization and propagation direction. The photonic band gap is therefore likely to be the true optical analog of the fundamental gap of a semiconductor. Since their invention in 1987, photonic crystals have triggered considerable interest because of their unusual optical properties. The unique properties of photonic crystals also led to their study being recognized as a new and major field of optoelectronics.
Macroporous silicon, with its high dielectric contrast, very high aspect ratios and full compatibility with the silicon microelectronic industry is an excellent model system for studying the optical properties of two-dimensional and even three-dimensional photonic crystals. Besides, macroporous silicon has been shown to have several unique uses in many other fields, like electronics, micromachining, gas sensing and biotechnology. Research into macroporous silicon is continuously growing, prompted by its enormous potential for applications.
The work presented in this thesis deals with two subjects: photonic band structure simulations and analysis of 2D photonic crystals, and the fabrication of macroporous silicon structures suitable for application as 2D infrared photonic crystals.
Since many potential applications of photonic crystals are based on their photonic band gaps, it is of interest to design photonic crystals with an absolute band gap that is as large as possible. In this thesis we describe a way to enlarge the absolute photonic band gap, showing the role that symmetry plays in designing optimal photonic structures. We have examined how reducing symmetry by inserting additional elements into the lattice unit cell or by changing the shape of the scatterers alters the dispersion behavior of the TMand TE-polarization modes in 2D photonic crystals. Our goal was to maximize the absolute PBG width by breaking the symmetry of the simple square and triangular lattices and thus to construct new structures, the so-called hybrid lattices. Using the FDTD method for photonic band structure calculations, we performed a detailed numerical analysis of the photonic dispersion relation in 2D hybrid lattices that consist of air holes drilled in silicon.
For square lattices, the symmetry reduction approach has been successfully applied to maximize the absolute PBG width. In the case of square lattices of circular air rods, the inclusion of an additional rod increases the absolute PBG threefold. For the case of square lattices of square air rods, the rotation of the rods plays a critical role in the opening of an absolute PBG. No absolute PBG was found if the square rods were not rotated. The size of the absolute PBG is improved most significantly by a combination of the inclusion of an additional rod and the rotation of square rods. Moreover, a new absolute PBG is generated that persists over a wide range of rotation angles and filling fractions, which are far from the closed-packed condition. This greatly favors the fabrication of photonic crystals.
The largest absolute PBG is the one for the triangular lattice of circular air rods.
Our results have shown that modifying the triangular structure by adding interstitial rods or using square rods (even though the rods are rotated) is not a good way of achieving a larger absolute PBG, at least for the special case of air/silicon structures. Adding more rods to the lattice unit cell cannot further enlarge the absolute PBG width.
We have made a detailed quantitative analysis of the absolute PBGs in 2D triangular and honeycomb lattices considering that there is an interfacial (shell) layer between the rods and the background dielectric matrix. This interfacial layer may be the unwanted result of the fabrication process itself or created intentionally. The properties of the photonic gaps are strongly affected by the thickness and the dielectric constant of the shell layer. The results of band structure simulations show that for structures consisting of air rods embedded in a dielectric background this layer reduces the absolute photonic gap.
For structures consisting of dielectric rods in air, however, an interfacial layer can yield larger photonic gaps if the dielectric constant of the layer is greater than that of the rods.
This provides further flexibility in the practical realization of such 2D photonic crystals.
For example, in certain cases we may not be able to obtain dielectric rods of the required diameter or of the particular material we need because of technological limitations.
However, we are enabled to grow the rods of materials with lower dielectric constants, for which a well-developed technology exists. The rods can then be covered with the required dielectric by deposition, thus achieving almost the same gap properties as those of the ideal shell-less structure.
We have developed an electrochemical etching set-up for fabricating 2D periodic structures based on macroporous silicon formation. We have also made a detailed study of how the electrochemical etching parameters influence the pore morphology. Straight and stable macropores can only be etched if all parameters of the etching process (doping level, HF concentration, etching current, anodic potential, temperature, etc.) are properly adjusted. The optimal conditions are only a very tiny part of the total parametric space, which requires a fine control of the process. For example, HF concentrations higher than 10 wt.%, which are commonly used for growing micro- and mesoporous films, are not suitable for growing deep, straight macropores. Relatively high anodic potentials (e.g. even higher than 2 V for our samples) inevitably enhance the formation of spiking breakdowntype pores on macropore walls. On the other hand, low anodic potentials (less than 1 V) usually lead to unstable pore growth with macropores that are partially or fully filled with microporous silicon.
Of all etching parameters the applied etching current is the most critical. Current densities greater than the critical current density Jps, which depends on the temperature and electrolyte concentration, will move the system into the electropolishing regime.
Controlling the etching current during the process is a key issue. Keeping the etching current constant was found not to be sufficient to grow deep, straight macropores. Two effects that influence the pore shape in depth were identified. First, the decrease in HF concentration towards the pore tips because of diffusional limitations leads to an increase of the pore diameter close to the tip. Second, the pore surface area increases for long anodization times, which leads to an increase in the dark current density and yields conical pores, the diameter of which decreases with depth. Increasing the etching current accordingly, which means to etch pores with the reverse conical shape is one of the methods to reduce the pore conicity. Performing the etching at lower temperatures and bubbling the electrolyte with nitrogen can reduce the dark current and produce less conical pores. Another effective way is to use appropriate surfactants. Surfactants are commonly used to prevent degeneration caused by bubbles sticking to the sample surface. Two surfactants of different types (nonionic TritonX-100 and anionic SDS) were tested. We found that the addition of nonionic surfactants increases the dark current contribution and thus enhances the formation of conical pores. On the other hand, the use of anionic surfactants considerably reduces the dark current and straight pores can be formed almost without difficulty. Highly uniform macropore arrays with different arrangements and dimensions were obtained by applying these "compensation" rules.
Finally, we have also presented some preliminary results of our work on novel applications of macroporous silicon. The structural features of the etched macropore arrays have been exploited to fabricate high-aspect-ratio silicon dioxide pillars, which may have applications in biotechnology as a 3D sensor platform for molecular recognition detections or as dense arrays of microsyringes for fluid delivery or precise chemical reaction stimulation. We have also fabricated a macroporous filter consisting of through-wafer pores and measured its optical characteristics. For light incidence parallel to the pores, a shortpass spectral behavior has been observed. The obtained results are only qualitative and suggest that further optimization of the etching process is needed in order to produce higher quality samples. We were also able to introduce periodic modulations of the pore diameter in depth and to fabricate ratchet-type macropore arrays, which have been envisioned for applications as ratchet devices for large-scale particle separation. We have shown that by a few post-etching steps the modulated macropore arrays can be converted into microstructures consisting of interconnected voids in all three dimensions. The technique used can be exploited for the fabrication of fully 3D photonic crystals.

Silicon carbide (SiC) nanostructures are appealing as non-toxic, water-stable and oxidation resistant nanomaterials. Owing to these unique properties, 3-dimensionally confined SiC nanostructures, namely SiC quantum dots (QDs) have found applications in bioimaging of living cells. Photoluminescence (PL) investigations however revealed that across the polytypes: 3C-, 4H- and 6H-SiC, excitation wavelength dependent PL is observed for larger sizes but deviate for sizes smaller than approximately 3 nm, thus exhibiting a dual-feature in the PL spectra. Additionally, nanostructures of varying polytypes and bandgaps exhibit strikingly similar PL emission centred at approximately 450 nm. At this wavelength, 3C-SiC emission is above bulk bandgap as expected of quantum size effects, but for 4H-SiC and 6H-SiC the emissions are below bandgap. 4H-SiC is a suitable polytype to study these effects. In this thesis, the hypotheses that mixed phases of polytypes or surface related defects obscuring the quantum confinement of 4H-SiC based nanostructure were investigated. Density functional theory (DFT) calculations within the Ab initio Modelling Programme (AIMPRO) were performed on OH-, F- and H-terminated 4H-SiC QDs with diameters in the range of 10 to 20 A° . The chosen surface terminations relate to the HF/ethanol electrolyte used in preparation of SiC QDs and the choice of size coincide with where deviation was observed in experiments. It was found that the absorption onset energies deviated from quantum confinement with -OH and -F terminations, but conform to the prediction when terminated with -H. The weak size-dependent absorption onsets for -OH and -F is due to surface states arising from lone pair orbitals that are spatially localised to the quantum dot surface where these terminations reside. On the other hand -H termination show strong size-dependent absorption onsets due to delocalisation of the electron wavefunction towards the quantum dot core assisting quantum confinement. It is predicted that the surface related states dominate up to sizes 25 and 27 °A for -F and -OH terminations respectively. As a result, the recombination mechanism would involve the interplay between quantum confinement and surface states affecting the resultant energy gap and the resulting PL. The PL would exhibit a dual-feature: excitation-wavelength independence for small sizes and excitationwavelength dependence for diameters larger than 3 nm as observed in the experiments. Mesoporous 4H-SiC was fabricated by anodic electrochemical etching in ethanoic HF electrolyte. The porous SiC suspended in ethanol exhibited three PL bands, those at wavelengths of 303 nm and 345 nm were rarely reported, above bulk bandgap and indicative of quantum confinement. The usually observed emission at 455 nm was below bulk bandgap. Dual-feature and below bandgap PL observed for wavelengths around 450 nm indicate that mesoporous 4H-SiC exhibited optical properties dictated by both quantum confinement (red-shifting with longer excitation wavelengths) and surface states (below bandgap). X-ray photoelectron spectroscopy provided evidence of -F, C=O and -COOH surface terminations that may contribute to these surface states. Raman scattering data exhibited a red-shift of 12.2 cm�1 and broadening in the lower frequency side of the longitudinal optical (LO) mode peak indicative of carrier depletion, surface phonons or phonon confinement as dimensions were reduced. The following ultrasonication process produced dimensions ranging from 16.9 5.5 down to 2.9 1.0 nm. The data from high resolution transmission electron microscopy (HRTEM) and X-ray diffraction (XRD) showed lattice spacing of 0.267 nm and peaks corresponding to the 4H-SiC polytype. No evidence of polytypic transformation from 4H-SiC to 3C-SiC resulting from ultrasonication was found in this work. Instead high crystallinity of 4H-SiC lattice was retained which suggested that the obscured quantum confinement may arise from surface effects rather than mixed polytypes. Thermal oxidation and subsequent HF dip of mesoporous 4H-SiC resulting in pore wall thinning and surface removal was undertaken. Cross sectional SEM analysis showed reductions in average pore wall thickness to (20.5 2.8) nm, (18.2 2.9) nm, (17.0 1.8) nm and (15.9 1.4) nm for 1, 3, 6 and 9 hours of oxidation respectively. Following ultrasonication, the PL and PL excitation (PLE) characterisation showed absorption/emission band centred at 290/325 nm which were above bandgap. The below bandgap emission centred at 455 nm was removed and is a significant finding. Surface removal by thermal oxidation and HF dip resulted in suppressed below bandgap PL but retained the above bandgap PL. The evidence strongly indicate that the dual-feature PL and below bandgap emission in 4H-SiC seen in experiments are surface related rather than due to polytypic transformation during ultrasonication.

Ces travaux de thèse portent sur la fabrication de via traversants conducteurs, brique technologique indispensable pour l’intégration des composants microélectroniques en 3 dimensions. Pour ce faire, une voie « tout-électrochimique » a été explorée en raison de son faible coût de fabrication par rapport aux techniques par voie chimique sèche. Ainsi, la gravure de macropores ordonnés traversants a été réalisée par anodisation du silicium en présence d’acide fluorhydrique puis leur remplissage de cuivre par dépôt électrochimique. L’objectif est de faire du silicium macroporeux une alternative crédible à la gravure sèche (DRIE) pour la structuration du silicium.Les conditions de gravure de matrices de macropores ordonnés traversants ont été étudiées à la fois dans des substrats silicium de type n et p faiblement dopés. La composition de l’électrolyte ainsi que le motif des matrices ont été optimisés afin de garantir la gravure de via traversants de forte densité et à facteur de forme élevé. Une fois gravés, les via traversant ont été remplis de cuivre. En optimisant ces paramètres une résistance minimale égale à 32 mΩ/via (soit 1,06 fois la résistivité théorique du cuivre à 20°C) a été mesurée
These thesis works deal with the achievement of Through Silicon Via (TSV) essential technological issue for microelectronic device 3D integration. For this purpose, we opted for a “full-electrochemical” way of TSV production because of lower fabrication costs as compared to dry etching and deposition techniques. Indeed, ordered through silicon macropores were carried out by silicon anodization in hydrofluoric acid-containing solution and then filled by copper electrochemical deposition. The main objective is to determine if the macroporous silicon arrays can be a viable alternative as Deep Reactive Ion Etching (DRIE).The etching parameters of through silicon macropore arrays were studied both in low-doped n- and p-type silicon. The electrolyte composition as well as the density of the initiation sites was optimized to enable the growth of high aspect ratio, high density through silicon ordered macropores. After silicon anodization, through via were filled with copper. By optimizing the copper deposition parameters (bath composition and applied potential), the resistance per via was measured equal to 32 mΩ (i.e. 1.06 times higher than the theoretical copper bulk resistivity)

Gas detection is of great importance in areas as diverse as industry, health or safety in domestic environments or public spaces, among others, and it is highly specific to each application. The detection method depends on factors such as the species of gas to be detected, concentration range, required resolution, sensitivity, specificity, response time, operating environment (temperature, humidity, interfering species, etc.), size and cost, among other considerations. Optical gas sensors are an attractive solution for gas detection. Most of them rely on molecular absorption and offer fast responses, minimal drift and are intrinsically reliable thanks to perform self-referenced measurements. Sensitivity and selectivity depend on the characteristics of the device. For example, laser-based gas sensors are highly selective with zero cross response to other gases and also with first-in-class sensitivity. The downside is that they are expensive. Non-dispersive infra-red (NDIR) sensors are a widespread alternative for cost-effective optical detection. They have inferior performances in terms of sensitivity and selectivity than laser-based sensors, but are two or three orders of magnitude less expensive. This thesis is dedicated to improving the selectivity and sensitivity of NDIR devices through the use of macroporous silicon technology. More specifically, it studies how photonic crystals manufactured by electrochemical etching can be used as narrow mid-infrared filters for gas detection purposes. That is, the photonic crystals are designed in such a way that only a small range of frequencies from an external source are transmitted while the surroundings are blocked. These filters are narrower than those available on the market and can be used to improve the selectivity and the sensitivity of NDIR devices as well as to reduce cross detection with other gases. In addition, the study shows how macroporous silicon photonic crystals can be heated to work as selective emitters. This can be used to reduce the complexity of the NDIR device while maintaining similar optical characteristics. Furthermore, it is proven that photonic molecules can be employed to perform dual detection in both transmission and emission, giving a new approach to self-referenced measurements. Conclusions of the work show that macroporous silicon technology is a versatile platform to provide solutions in the mid-infrared range for developing compact, sensitive and selective optical gas sensing.
La detecció de gasos és de gran importància en àrees tan diverses com la indústria, la salut o la seguretat en entorns domèstics o espais públics, entre d'altres, i és altament específica per a cada aplicació. El mètode de detecció a utilitzar depèn de factors com ara el gas a detectar, el rang de concentració, la resolució requerida, la sensibilitat, l'especificitat, el temps de resposta, l'entorn operatiu (temperatura, humitat, espècies interferents, etc. .), la mida i el cost, entre altres consideracions. Els sensors òptics de gas són una solució atractiva per a la detecció de gas. La majoria d'ells es basen en l'absorció molecular i ofereixen respostes ràpides, deriva mínima i són intrínsecament fiables gràcies a la realització de mesures auto-referenciades. La sensibilitat i la selectivitat depenen de les característiques del dispositiu. Per exemple, els sensors de gas basats en tecnologia làser són altament selectius, no presenten resposta creuada a altres gasos i són altament sensibles. El desavantatge és que són cars. Els sensors d'infrarojos no dispersius (NDIR) són una alternativa molt estesa per a la detecció òptica de baix cost. Tenen un rendiment inferior en termes de sensibilitat i selectivitat que els sensors basats en làser, però són dos o tres ordres de magnitud més barats. Aquesta tesi està dedicada a millorar la selectivitat i la sensibilitat dels dispositius NDIR mitjançant la tecnologia de silici macroporós. Més específicament, estudia com els cristalls fotònics fabricats mitjançant el gravat electroquímic poden ser usats com a filtres estrets d'infraroig mitjà per a la detecció de gasos. És a dir, els cristalls fotònics estan dissenyats de tal manera que només un petit rang de freqüències d'una font externa es transmet mentre que els voltants estan bloquejats. Aquests filtres són més estrets que els disponibles en el mercat i poden utilitzar-se per millorar la selectivitat i la sensibilitat dels dispositius NDIR, així com per reduir la detecció creuada amb altres gasos. A més, l'estudi mostra com els cristalls fotònics de silici macroporós poden funcionar com a emissors selectius si són escalfats. Això pot ser usat per reduir la complexitat dels dispositius NDIR alhora que es mantenen característiques òptiques similars. A més, s'ha demostrat que les molècules fotòniques poden emprar-se per realitzar una detecció dual tant en la transmissió com en l'emissió, donant un nou enfocament a les mesures auto-referenciades. Les conclusions del treball mostren que la tecnologia de silici macroporós és una plataforma versàtil que proporciona solucions en el rang d'infraroig mitjà per al desenvolupament de sensors de gas òptics compactes, sensibles i selectius.

While silicon represents the dominant material in the semiconductor industry, the continuous improvement in the performance of Si based devices is reaching its upper bound due to the approaching of insuperable physical limitations intrinsic to Si, which requires the introduction of new semiconductor materials and the development of new assembly techniques to guarantee the future performance improvement and reduction in fabrication costs. The integration of high-quality germanium epilayers on Si substrates has received great attention from the semiconductor community due to the chance to extend the range of performance offered by Si-based technology by taking advantage of both the superior properties of Ge such as a higher carrier mobility, a lattice constant close to that of GaAs which enables III-V epitaxy and a quasi-direct bandgap, and of the possibility of strain and bandgap engineering offered by the formation of a heterojunction. To overcome the 4.2% lattice constant mismatch existing between Ge and Si which hamper the direct integration approach, this thesis investigates a novel technique for the realization of high-quality Ge on Si virtual substrates (VSs), consisting in the introduction of a porous silicon (pSi) buffer layer in between Ge and Si. pSi is a versatile, self-assembled, nanomaterial which can be realized at very high growth rates through electrochemical etching of Si. Thanks to its reduced Young’s and shear moduli pSi can deform during epitaxy, potentially alleviating part of the lattice mismatch between Ge and Si and reducing the density of misfit dislocations and associated threading segments necessary for complete Ge relaxation. Together with the very high throughput of the anodization process, other fundamental advantages of the proposed approach are its low cost, its simple scalability to large area Si substrates and the possibility to lift-off the grown epilayers from the starting substrates, giving Ge on pSi VSs the possibility to outperform other existing techniques for Ge integration on Si. During the course of this work, several Ge on pSi VSs have been grown through low energy plasma enhanced chemical vapor deposition (LEPECVD) technique, and the resulting crystalline quality has been compared to that of Ge on Si VSs. Using X-ray diffraction techniques, together with electron microscopy analysis and selective etching techniques, it will be shown how the main physical parameters of pSi buffers affect the crystalline quality of Ge heteroepilayers. Finally, it will be demonstrated that strong threading dislocation reduction is possible in Ge grown on low porosity pSi buffers compared to Ge on bulk Si, at parity of experimental conditions, and the main mechanisms responsible for crystalline quality improvement in Ge grown on pSi will be uncovered.

Ces travaux portent sur la localisation de la gravure du silicium au moyen de masques polymères résistants à l'acide fluorhydrique, indispensable à la formation du silicium poreux par gravure électrochimique ou par gravure chimique catalysée par métal (MaCE). Pour ce faire, nous avons exploré une méthode d'élaboration des masques « tout-chimie » : la séparation de phases induite par évaporation du solvant (SPIES) dans un mélange de polymères déposé par spin-coating suivie d'une extraction sélective. L'objectif est de réaliser des masques de gravure du silicium sans avoir recours à une étape plasma et présentant des ouvertures sub-micrométriques. Cette méthode bien que rapide et facile à mettre en œuvre, met en compétition des phénomènes thermodynamiques et cinétiques complexes déterminant la morphologie finale (i.e. après séchage) des films polymères. La compréhension des mécanismes mis en jeu et le choix judicieux des paramètres expérimentaux ont permis de réaliser des matrices polymères perforées régulièrement mais également des domaines discrets avec des dimensions sub-micrométriques. Des gravures électrochimiques du silicium ont été réalisées à travers ces masques en optimisant la densité de courant et la composition de l'électrolyte. Le caractère protecteur des masques a été évalué et a mis en évidence le fait que la séparation de phases secondaire inhérente à la méthode de structuration des masques, peut créer des chemins de percolation que peut emprunter l'électrolyte et le courant électrique et réduire l'imperméabilité du masque. Bien que le caractère protecteur des masques soit limité pour la gravure électrochimique, ils se révèlent prometteurs pour la formation de silicium poreux par MaCE. Ce procédé a été mis en œuvre par argent et par or au travers des masques et a permis de structurer des micro-colonnes et des micro-piliers ouvrant la méthode SPIES à de nouvelles applications
This work deals with the localisation of silicon etching by using polymer masks resistant to hydrofluoric acid, which is indispensable for the formation of porous silicon by electrochemical etching or metal-assisted chemical etching (MaCE). For this purpose, we explored for producing “all-chemical” masks: solvent evaporation-induced phases separation (SEIPS) in a spin-coated polymer blend followed by selective extraction. The main objective is to produce silicon etching masks without the need for a plasma step with sub-micrometre apertures. Although this method is quick and easy to implement, it involves complex thermodynamic and kinetic phenomena in competition that determine the final morphology (i.e. after drying) of the polymer films. By understanding the mechanisms involved and choosing the right experimental parameters, we were able to produce regularly perforated polymer matrices as well as discrete domains with sub-micrometric dimensions. Electrochemical etchings of silicon were performed through these masks by optimising current density and electrolyte composition. The protective property was assessed, highlighting the fact that the secondary phase separation, inherent in the mask structuring method, can create percolation paths. The electrolyte and the electric current can flow through those percolation paths, reducing the impermeability of the mask. Although the protective property of the masks is limited for electrochemical etching, they are promising for the formation of porous silicon by MaCE. The process has been implemented using silver and gold through the masks and enables micro-columns and micro-pillars to be structured, opening up the SPIES method new applications

Ces travaux de thèse ont pour but l’évaluation et le développement de briques technologiques en silicium poreux répondant à la problématique de l’intégration monolithique 3D rattachée au concept du “more than Moore” : d’une part l’intégration sur silicium de composants passifs RF, d’autre part, la réalisation de chemins traversants d’interconnexion à fort facteur d’aspect par voie électrochimique. Dans un premier temps, différents substrats mixtes silicium / silicium poreux sont réalisés. Des inductances en cuivre, réalisées sur un substrat mésoporeux de 200 µm de profondeur et de porosité proche de 60%, atteignent des facteurs de qualité à 20 GHz jusqu’à 55% supérieurs à ceux mesurés sur silicium massif. Une perspective d’industrialisation de ce type d’application est à l’étude dans le cadre d’une thèse CIFRE. La gravure de matrices de pores à fort facteur d’aspect, bien qu’encore difficilement localisable en termes de qualité de périphérie, fait d’autre part l’objet de développements, notamment pour la fabrication de condensateurs à haute densité capacitive et de contacts d’interconnexions en cuivre
Those thesis works deal with the evaluation and the development of porous silicon technological step in order to answer some of the monolithic integration challenges bring by the “more than Moore” problematic in microelectronics industry: on one hand, the integration on silicon of passive RF devices, on the other hand, realization by electrochemical etching of through silicon via. In a first time, several mixed porous silicon / silicon substrat are realized. Copper inductors, realized on 200 µm thick and 60% porosity mesoporous layer, show a quality factor superior to 55% to the one obtained on massive silicon. Industrialization perspectives are on the line via a CIFRE PhD convention. In a second time, several electrochemical etching process are evaluated. Among them, high aspect ratio macropore array etching, although poorly localizable, allows many perspectives: copper via and high density capacitor

Die vorliegende Arbeit befasst sich mit der grundlegenden Untersuchung von Reaktionsmustern kristalliner Si-Oberflächen in HF-basierten Lösungen. Ausgehend von den industriell genutzten HF-HNO3-H2O-Gemischen wurden wisher wenig untersuchte HF/HNO3-Konzentrationsverhältnisse, die durch gelöste Stickoxide bedingten Folgereaktionen sowie der PH-Wert als Steuerparameter zur Aufarbeitung feinkörniger Si-Rohstoffe (Korngröße ≤ 0,5 mm) identifiziert. Die in diesem Kontext zentrale Rolle der NO+-Ionen wurde durch Untersuchung der spezifischen Reaktionsmuster an kristallinen as-cut und hydrophobierten Si-Oberflächen sowie bei Umsetzungen mit Oligosilanen als Modellverbindungen bestätigt. Die aus den umfassenden analytischen Daten (FT-IR-, Raman-, NMR-Spektroskopie, IC, REM-EDX, AFM) gewonnenen Erkenntnisse liefern einen wichtigen Beitrag zum Verständnis nasschemischer Halbleiterätzprozesse und erschließen neue Anwendungsfelder.

Die vorliegende Arbeit befasst sich mit der grundlegenden Untersuchung von Reaktionsmustern kristalliner Si-Oberflächen in HF-basierten Lösungen. Ausgehend von den industriell genutzten HF-HNO3-H2O-Gemischen wurden wisher wenig untersuchte HF/HNO3-Konzentrationsverhältnisse, die durch gelöste Stickoxide bedingten Folgereaktionen sowie der PH-Wert als Steuerparameter zur Aufarbeitung feinkörniger Si-Rohstoffe (Korngröße ≤ 0,5 mm) identifiziert. Die in diesem Kontext zentrale Rolle der NO+-Ionen wurde durch Untersuchung der spezifischen Reaktionsmuster an kristallinen as-cut und hydrophobierten Si-Oberflächen sowie bei Umsetzungen mit Oligosilanen als Modellverbindungen bestätigt. Die aus den umfassenden analytischen Daten (FT-IR-, Raman-, NMR-Spektroskopie, IC, REM-EDX, AFM) gewonnenen Erkenntnisse liefern einen wichtigen Beitrag zum Verständnis nasschemischer Halbleiterätzprozesse und erschließen neue Anwendungsfelder.

Les nanofils de silicium (SiNWs) attirent l’attention particulière en raison de leurs applications thermoélectriques prometteuses. La faible conductivité thermique et les propriétés électriques proches du Si massif en font un nanomatériau thermoélectrique idéal dans le concept de "verre à phonons - cristal à électrons". Théoriquement, les valeurs du facteur de mérite thermoélectrique (ZT) pour SiNW peuvent atteindre 3 à la température ambiante. ZT = 0,7 a été déjà obtenu expérimentalement pour des SiNW individuels, ce qui est proche de ZT pour les chalcogénures de bismuth (ZT = 0,8 -1,0) qui sont couramment utilisés. De point de vue pratique, il faut savoir fabriquer des réseaux de SiNWs à faible coût. Dans cette thèse, nous avons cherché: (i) à adapter des technologies existantes pour la fabrication des SiNWs fortement dopés, (ii) à développer des méthodes de caractérisation sans contact et non destructive des propriétés électriques et thermiques de réseaux de SiNWs, (iii) à fabriquer et caractériser des réseaux de SiNWs à haute conductivité électrique et faible conductivité thermique. Les réseaux des SiNW ayant la morphologie et le niveau de dopage nécessaires pour obtenir un ZT maximal ont été fabriquées par gravure chimique assistée par métaux de substrats de silicium. Une procédure de dopage post-fabrication a été développée en utilisant la diffusion thermique d’atomes de dopant à partir de solutions de dopage (via un dépôt spin-on). En particulier, les réseaux de nanofils de silicium ayant un diamètre typique de 100 nm, une longueur de 10 mm, une morphologie de type "cœur cristallin/ surface rugueuse" et un niveau de dopage de 10 20 cm
Silicon nanowires (SiNWs) attract growing attention in view of their promising thermoelectric applications. Low thermal conductivity and bulk-like electrical properties make them a perfect candidate as a thermoelectric material in framework of the concept “phonon-glass/ electroncrystal”. Theoretically, the values of figure of merit (ZT) for SiNWs as high as three can be achieved at room temperature, and experimentally ZT = 0.7 were already observed for individual SiNWs, which is close to ZT for commonly used bismuth chalcogenides (ZT = 0.8-1.0). For practical application of SiNWs, the low-cost fabrication methods for SiNWs arrays with high ZT should be achieved. In this thesis we aimed: (i) to adapt available semiconductor technology for fabrication of highly-doped SiNWs arrays, (ii) to develop contactless methods for non-destructive characterization of electrical and thermal properties of the SiNWs arrays, (iii) to fabricate and characterize SiNWs arrays with high electrical and low thermal conductivities. The arrays of SiNWs with the morphology and doping level necessary for maximum ZT were fabricated using metal-assisted chemical etching of silicon wafers and post-fabrication doping procedure, which consisted of the thermal diffusion of dopant atoms from spin-on dopant solutions. In particular, the arrays of silicon nanowires with a typical diameter of 100 nm, length of 10 mm, bulk core/rough surface morphology and doping level of 1020 cm

L’objectif de ce travail est d’utiliser les nanoparticules (NPs) de nanosondes fluorescentes de température en particulier dans les films lubrifiants. Le développement de ces nanosondes nécessite la détermination de leurs sensibilités thermiques afin de pouvoir sélectionner les NPs les plus prometteuses. Pour atteindre cet objectif, nous avons présenté deux méthodes d’élaboration utilisées pour la synthèse des nanostructures à base de SiC-3C, la méthode d’anodisation électrochimique et la méthode d’attaque chimique. Dans le premier cas, les analyses FTIR,RAMAN et MET des NPs finales ont montré que la nature chimique de ces NPs est majoritairement formée de carbone graphitique. L’étude détaillée de la photoluminescence de ces NPs a montré que le processus d’émission dépend de la chimie de surface des NPs, du milieu de dispersion et de sa viscosité, de la concentration des suspensions et de la température du milieu. Pour la deuxième famille de NP de SiC, les analyses cohérentes MET, DLS et PL ont montrées une taille moyenne de 1.8 nm de diamètre avec une dispersion de ±0.5nm. Le rendement quantique externe de ces NPs est de l’ordre de 4%. Les NPs dispersées dans l’éthanol, n’ont pas montré une dépendance à la température exploitable pour notre application. Par contre, les NPs de SiC produites par cette voie, étant donné la distribution en taille resserrée et le rendement quantique « honorable » pour un matériau à gap indirect, sont prometteuses pour des applications comme luminophores en particulier pour la biologie grâce à la non toxicité du SiC. Dans le cas des NPs de Si, nous avons également étudié deux types différents de NPs. Il s’agit de : (i) NPs obtenues par anodisation électrochimique et fonctionnalisées par des groupements alkyls (décène, 1-octadécène). Nous avons mis en évidence pour la première fois une très importante variation de l’énergie d’émission dEg/dT avec la température de type red-shift entre 300 et 400K. Les mesures de(T) conduisent à une sensibilité thermique de 0.75%/°C tout à fait intéressante par rapport aux NPs II-VI. De plus il a été montré que la durée de vie mesurée n’est pas fonction de la concentration. (ii) NPs obtenue par voie humide et fonctionnalisées par le n-butyl. Pour ce type de NPs nous avons mis pour la première fois en évidence un comportement de type blue-shift pour dEg/dT de l’ordre de -0.75 meV/K dans le squalane. Pour ces NPs, la sensibilité thermique pour la durée de vie de 0.2%°C est inférieure à celle des NPs de type (i) mais largement supérieure à celle des NPs de CdSe de 4 nm (0.08%/°C). La quantification de cette la sensibilité à la température par la position du pic d’émission dEg/dT et de la durée de vie nous permet d’envisager la conception de nanosondes de température basée sur les NPs de Si avec comme recommandations l’utilisation de NPs obtenues par anodisation électrochimique et de la durée de vie comme indicateur des variations en température
The goal of this study is the use of Si and SiC nanoparticles (NPs) as fluorescent temperature nanoprobes particularly in lubricating films. The development of these nanoprobes requires the determination of their thermal sensitivity in order to select the best prospects NPs. To achieve this goal, we presented two preparation methods used for the synthesis of 3C-SiC based nanostructures : (i) anodic etching method and (ii) chemical etching method. In the first case, the FTIR, Raman and TEM analysis of final NPs showed that the chemical nature of these NPs is formed predominantly of graphitic carbon. The detailed photoluminescence study of these NPs showed that the emission process depends on the surface chemistry of the NPs, the dispersion medium and its viscosity, the suspension concentration and temperature of the environment.. In the second case, coherent TEM, DLS and PL analyzes showed an average size of 1.8 nm in diameter with a dispersion of ±0.5 nm. The external quantum efficiency of these NPs is 4%. NPs dispersed in ethanol, did not show an exploitable fluorescence dependence on temperature for our application. On the other hand, 3C-SiC NPs produced by this way, given the narrow size distribution and the reasonably high quantum yield for an indirect bandgap material, are promising for applications such as luminophores in particular in the biology field thanks to nontoxicity of SiC. In the case of Si we studied also two different types of NPs. (i) NPs obtained by anodic etching and functionalized by alkyl groups (decene, octadecene). We have demonstrated for the first time an important red-shift in the emission energy dEg/dT with temperature from 300 to 400K. The PL lifetime measurement(T) lead to a thermal sensitivity of 0.75% /°C very interesting compared to II-VI NPs. Furthermore it has been shown that t is not depending on the concentration. (ii) NPs obtained by wet-chemical process and functionalized with n-butyl. For this type of NPs we have identified for the first time a blue-shift behavior of dEg dT in the order of -0.75 meV/K in squalane. The thermal sensitivity for the PL lifetime of these NPs is 0.2%/°C, which is lower than that of NPs obtained by anodic etching method, but much greater than that of CdSe NPs with 4 nm of diameter (0.08%/°C). Quantification of the temperature sensitivity by the position of emission peak dEg/dT and the PL lifetime dτ/dT allows us to consider the realization of temperature nanoprobes based on Si NPs with recommendations to use Si NPs obtained by anodic etching method and PL lifetime as an indicator of temperature changes

This thesis presents methods for fabricating MicroElectroMechanical System (MEMS) actuators and high-flow gas microvalves using wafer-level integration of Shape Memory Alloys (SMAs) in the form of wires and sheets. The work output per volume of SMA actuators exceeds that of other microactuation mechanisms, such as electrostatic, magnetic and piezoelectric actuation, by more than an order of magnitude, making SMA actuators highly promising for applications requiring high forces and large displacements. The use of SMAs in MEMS has so far been limited, partially due to a lack of cost efficient and reliable wafer-level integration approaches. This thesis presents new methods for wafer-level integration of nickel-titanium SMA sheets and wires. For SMA sheets, a technique for the integration of patterned SMA sheets to silicon wafers using gold-silicon eutectic bonding is demonstrated. A method for selective release of gold-silicon eutectically bonded microstructures by localized electrochemical etching, is also presented. For SMA wires, alignment and placement of NiTi wires is demonstrated forboth a manual approach, using specially built wire frame tools, and a semiautomatic approach, using a commercially available wire bonder. Methods for fixing wires to wafers using either polymers, nickel electroplating or mechanical silicon clamps are also shown. Nickel electroplating offers the most promising permanent fixing technique, since both a strong mechanical and good electrical connection to the wire is achieved during the same process step. Resistively heated microactuators are also fabricated by integrating prestrained SMA wires onto silicon cantilevers. These microactuators exhibit displacements that are among the highest yet reported. The actuators also feature a relatively low power consumption and high reliability during longterm cycling. New designs for gas microvalves are presented and valves using both SMA sheets and SMA wires for actuation are fabricated. The SMA-sheet microvalve exhibits a pneumatic performance per footprint area, three times higher than that of previous microvalves. The SMA-wire-actuated microvalve also allows control of high gas flows and in addition, offers benefits of lowvoltage actuation and low overall power consumption.
QC 20120514

Le domaine du photovoltaïque en couches minces s’attache à réduire le coût de l’énergie photovoltaïque, en réduisant considérablement la quantité de matières premières utilisées. Dans le cas du silicium cristallin en couches minces, la réduction de l’épaisseur de la cellule s’accompagne d’une baisse drastique de l’absorption, notamment pour les plus fortes longueurs d’onde. Nombreuses sont les techniques aujourd’hui mises en œuvre pour lutter contre cette baisse de performance, dont l’utilisation des effets plasmoniques induits par des nanostructures métalliques qui permettent un piégeage de la lumière accru dans la couche absorbante. Dans ces travaux, nous étudions l’influence de nanostructures d’argent organisées suivant un réseau périodique sur l’absorption d’une couche de silicium. Ces travaux s’articulent autour de deux axes majeurs. L’influence de ces effets plasmoniques sur l’absorption est d’abord mise en évidence à travers différentes simulations numériques réalisées par la méthode FDTD. Nous étudions ainsi les cas de réseaux périodiques finis et infinis de nanostructures d’argent situés sur la face arrière d’une couche mince de silicium. En variant les paramètres du réseau, nous montrons que l’absorption au sein du silicium peut être améliorée dans le proche infrarouge, sur une large plage de longueurs d’onde. Le second volet de la thèse concerne la réalisation des structures modélisées. Pour cela, deux voies de fabrication ont été explorées et développées. Pour chacune d’entre elles, trois briques élémentaires ont été identifiées : (i) définition du futur motif du réseau grâce à un masque, (ii) réalisation de pores dans le silicium et (iii) remplissage des pores par de l’argent pour former le réseau métallique. La première voie de fabrication développée fait appel à un masque d’alumine, réalisé par l’anodisation électrochimique d’une couche d’aluminium, pour définir les dimensions du réseau métallique. Une gravure chimique assistée par un métal est ensuite utilisée pour former les pores, qui seront alors comblés grâce à des dépôts d’argent par voie humide. La seconde voie de fabrication utilise un masque réalisé par lithographie holographique, une gravure des pores par RIE et un remplissage des pores par dépôt d’argent electroless. Les substrats plasmoniques fabriqués sont caractérisés optiquement, au moyen d’une sphère intégrante, par des mesures de transmission, réflexion et absorption. Pour tous les substrats plasmoniques caractérisés, les mesures optiques montrent une baisse de la réflexion et de la transmission et une hausse de l’absorption pour les plus grandes longueurs d’onde
Thin-film photovoltaics focus on lowering the cost reduction of photovoltaic energy through the significant reduction of raw materials used. In the case of thin-films crystalline silicon, the reduction of the thickness of the cell is linked to a drastic decrease of the absorption, particularly for the higher wavelengths. This decrease of the absorption can be fought through the use of several different light trapping methods, and the use of plasmonic effects induced by metallic nanostructures is one of them. In this work, we study the influence of a periodic array of silver nanostructures on the absorption of a silicon layer. This work is decomposed into two main axes. First, the influence of the plasmonic effects on the silicon absorption is highlighted through different numerical simulations performed by the FDTD method. Both finite and infinite arrays of silver nanostructures, located at the rear side of a thin silicon layer, are studied. By varying the parameters of the array, we show that the silicon absorption can be improved in the near infrared spectral region, over a wide range of wavelengths. The second part of the thesis is dedicated to the fabrication of such modeled structures. Two different approaches have been explored and developed inside the lab. For each of these two strategies, three major building blocks have been identified: (i) definition of the future array pattern through a mask, (ii) etching of the pattern in the silicon layer and (iii) filling of the pores with silver in order to form the metallic array of nanostructures. In the first fabrication method, an anodic alumina mask, produced by the electrochemical anodization of an aluminium layer, is used in order to define the dimensions of the metallic array. A metal assisted chemical etching is then performed to produce the pores inside the silicon, which will then be filled with silver through a wet chemical process. The second fabrication method developed involves the use of holographic lithography to produce the mask, the pores in silicon are formed by reactive ion etching and they are filled during an electroless silver deposition step. The fabricated plasmonic substrates are optically characterized using an integrating sphere, and transmission, reflection and absorption are measured. All the characterized plasmonic substrates shown a decrease of their reflection and transmission and an absorption enhancement at the largest wavelengths

[ES] El desarrollo de los biosensores está permitiendo llevar a cabo análisis bioquímicos cada vez más rápidos, de manera mucho más sencilla y utilizando una menor cantidad de muestra. Esto está dando lugar a aplicaciones en las que se monitorizan parámetros de manera continua y autónoma, aumentando la eficiencia y reduciendo los costes. El tema principal de esta Tesis ha sido el desarrollo y la evaluación de biosensores que se basan en técnicas de transducción óptica, fabricados en silicio poroso, un material nanoestructurado que puede llegar a alcanzar una gran sensibilidad. El trabajo ha consistido en el estudio de la fabricación y la caracterización de membranas de silicio poroso obtenidas a partir de substratos tipo p de baja resistividad. Para ello se ha desarrollado un modelo matemático realista que permite simular el comportamiento del transductor y calcular sus parámetros experimentales. Gracias a esto, se han estudiado propiedades del material como el efecto térmico, llevando a caracterizar el efecto termo-óptico del silicio poroso en el rango infrarrojo del espectro. Además, se ha analizado la infiltración de la muestra en el transductor con el objetivo de mejorar su funcionamiento. Por este motivo, se han examinado diferentes morfologías de poros y se ha implementado un flujo activo durante el sensado, en el cual la sustancia a analizar fluye a través de la membrana porosa, resolviendo problemas de rellenado del sensor y mezclado con otras sustancias.
[CAT] El desenvolupament dels biosensors està permetent realitzar anàlisis bioquímics cada vegada més ràpids, de manera molt més senzilla i utilitzant una menor quantitat de mostra. Això està donant lloc a aplicacions en les quals es monitoritzen paràmetres de manera contínua i autònoma, augmentant l'eficiència i reduint els costos. El tema principal d'aquesta Tesis ha sigut el desenvolupament i l'avaluació de biosensors basats en tècniques de transducció òptica, fabricats en silici porós, un material nanoestructurat que pot arribar a aconseguir una gran sensibilitat. El treball ha consistit en l'estudi de la fabricació i la caracterització de membranes de silici porós obtingudes a partir de substrats tipus p de baixa resistivitat. Per a fer-ho, s'ha desenvolupat un model matemàtic realista que permet simular el comportament del transductor i calcular els seus paràmetres experimentals. Gràcies a això, s'han estudiat propietats del material com l'efecte tèrmic, el que ha permés caracteritzar l'efecte termo-òptic del silici porós en el rang infraroig de l'espectre. A més, s'ha analitzat la infiltració de la mostra en el transductor amb l'objectiu de millorar el seu funcionament. Per aquest motiu, s'han examinat diferents morfologies de porus i s'ha implementat un flux actiu durant el sensat, en el qual la substància a analitzar fluïx a través de la membrana porosa, resolent problemes d'ompliment del sensor i mesclat amb altres substàncies.
[EN] The development of biosensors is leading to faster and simpler analyses of biochemical samples, using them in lower quantities. Over the last years, these advances have allowed the emergence of applications where parameters can be monitored continuously and autonomously, increasing the efficiency and reducing the costs. This Thesis has focused on the development and evaluation of biosensors based on optical transducers, which are fabricated with porous silicon, a nanostructured material that is able to reach a high sensitivity. In this work, the fabrication and characterization of porous silicon membranes using heavily doped p-type silicon wafers have been studied. A realistic mathematical model has been developed in order to simulate the transducer's behavior and calculate the experimental parameters. This has led to the study of physical properties such as the thermal effect, where we were able to characterize the thermo-optic coefficient in the near-infrared range. Moreover, the penetration of the sample into the structure has been analyzed. For this purpose, several pore morphologies were examined and an active flow has been implemented during the sensing experiments, where the substance of interest flows through the porous membrane, to solve problems such as the partial filling of the sensor or the mixture of different substances during the experiments.
Martín Sánchez, D. (2019). Desarrollo de biosensores fotónicos basados en membranas de silicio poroso [Tesis doctoral no publicada]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/125695
TESIS

Les transducteurs ultrasonores capacitifs micro-usinés (CMUT) représentent aujourd’hui une réelle alternative aux technologies piézoélectriques dans le domaine de l’imagerie échographique médicale. Au cours des années, les procédés de fabrication des transducteurs se sont enrichis en vue d’améliorer leurs performances. A contrario le choix du substrat, généralement en silicium, a été peu étudié. Il est cependant reconnu que le support contribue à la signature acoustique du dispositif ultrasonore. L’objectif de ces travaux de thèse a été d’intégrer une couche de silicium poreux afin d’absorber une partie des ondes élastiques qui se propagent dans le substrat et interfèrent avec le signal acoustique émis. Nous montrons alors qu’il été possible de réaliser une couche de silicium poreux en face arrière de composants, sur plaquettes 6 pouces, sans dégrader leurs performances. Finalement, par l’intermédiaire de caractérisations acoustiques et des signatures impulsionnelles des transducteurs, nous révélons le potentiel prometteur de ce matériaux pour la réalisation de milieu arrière atténuant dédié à la transduction ultrasonore
Capacitive micromachined ultrasonic transducers (CMUT) have emerged as a potential alternative to traditional piezoelectric transducers for ultrasound imaging. Along the years, CMUT processes have been evolved to enhance the device performances. In the meantime, no particular attention was paid on the silicon substrate, even if it is well-known that it could contribute to the transducer efficiency. The aim of this PhD thesis was to use porous silicon as a backing material for ultrasonic transducers to absorb a piece of the acoustic wave propagating in the substrate and which induce crosstalks in the acoustic signal. We show that porous silicon layer can be obtained on the rear side of already processed wafers without any damage on the performances of capacitive micromachined ultrasonic transducers. Finally, by means of acoustic characterizations and the transducer electroacoustic responses, we reveal the potential interest of porous silicon as backing material for ultrasonic transducers

De nos jours, la modification de surface fait l'objet d'attentions particulières en raison de sa variété d'applications dans divers domaines. Dans ce contexte, l'objectif de cette thèse a été de traiter de la modification localisée de surfaces dans des conditions douces en utilisant le microscope électrochimique à balayage (SECM). En tant que preuve de concept pour l'écriture directe, différentes stratégies ont été menées pour la modification de surface par gravure de matériaux, greffage de couche organique et modification de la structure chimique surfacique. Des surfaces de verre et d'or ont été les principaux substrats qui ont été modifiés du fait de leur large utilisation notamment dans les nanotechnologies. Cette thèse est présentée sous forme de quatre chapitres et le premier est consacré à la technique SECM ainsi qu'à la modification de surface en général. Les trois autres parties concernent le travail effectué pour valider le concept d'écriture directe. Dans la première partie, une matrice de silice synthétisée par voie sol-gel et dopée avec un ion métallique (cuivre ou or) est utilisée comme matériau d'écriture à l'aide d'une sonde locale (ultramicroéléctrode). Le SECM est utilisé en mode de rétroaction avec des médiateurs tels que viologène de méthyle et le p-benzoquinone. Le diamètre de l'ultramicroélectrode (UME) et la durée d'hydrolyse ont été des facteurs pris en compte pour étudier l'effet sur la taille des plots métalliques électrogénérés. Dans la deuxième partie, la gravure par voie humide localisée de la surface de l'or a été réalisée en utilisant le SECM opérant dans un électrolyte à base de diméthylsulfoxide chargé avec de l'iode. Dans cette méthode, une UME est positionnée (à une distance connue) à proximité de la surface d'or pour générer électrochimiquement l'ion triiodure à la pointe de l'UME de platine, agissant comme oxydant à la surface d'or. La troisième partie comprend deux travaux expérimentaux différents mais complémentaires. Le premier porte sur la réduction électrochimique sur électrode d'or d'un sel de diazonium préparé à partir de l'éthylènediamine, une molécule aliphatique. Pour la première fois, la fonctionnalisation covalente sur or d'un sel de diazonium est démontrée via la diazotation d'un groupe amino de l'éthylènediamine. Dans la seconde partie, un substrat de verre a été greffé par un film à base de 3-aminopropyle silane qui a été réalisée par un procédé sol-gel. Ensuite, la lame de verre modifiée a été fonctionnalisée avec du glutaraldéhyde pour greffer la tyrosinase. Enfin, le mode de réaction du SECM a également été utilisé pour vérifier l’activité catalytique de cette enzyme. La pointe de l’UME est positionnée à proximité de la surface modifiée par l’enzyme afin de réaliser une mesure de courant de l’activité enzymatique à partir d’un balayage horizontal dans le plan x-y
Nowadays, the modification of surfaces has drawn more attention due to its variety of applications in various domains. Therefore, the purpose of this thesis deals with the localized modification of surfaces in mild condition by using the scanning electrochemical microscope (SECM) instrument. As a proof of concept for direct writing, different strategies have been used for surface modifications through removing surface materials, grafted organic layer and changing the chemical structure of the surface. Gold wafer and glass surfaces were the main substrates which have been modified since these materials are very used especially in nanotechnologies. This dissertation is conducted in four chapters and the first one focuses on SECM technique and surface modifications in general. The three other parts concern the work performed to validate the concept of direct writing. In the first part, metal ion (copper and gold)-doped silica matrices have been prepared by the well-known sol-gel method. Copper and gold metallic particles are produced locally by using the SECM in feedback mode with mediators such as methyl viologen and p-benzoquinone. The diameter of ultramicroelectrode (UME) tip and hydrolysis period were factors taken into account to study the effect on the size of electrogenerated metallic spots. In the second part, the localized wet etching of gold surface has been achieved by using SECM where a dimethylsulfoxide-based electrolyte charged with iodine is used. In this method an UME probe is positioned (at a known distance) close to the gold surface. Friendly environment method was used as etching process to generate electrochemically triiodide ion at the platinum UME tip, acting as an oxidant for gold surface. he third part includes two different experimental works. The first one covers the electrochemical reduction on gold electrode of diazonium salt prepared from ethylenediamine, an aliphatic diamine molecule. For the first time, the covalent functionalization on gold of a diazonium salt is demonstrated, and required diazotization of one amine group from ethylenediamine. In the second work, glass substrate was grafted by 3-aminopropyl silane film which was performed by sol-gel method. Then the modified-glass slide was functionalized by glutaraldehyde solution in order to immobilize tyrosinase molecules. Finally, the feedback mode of SECM has also been used to monitor the catalytic activity of tyrosinase. The tip of ultramicroelectrode was positioned close to the enzyme-modified surface and was scanned horizontally in x-y plane while measuring current from re-generated mediator molecules was carried out

A study of potentiostatic and galvanostatic electrochemical etching of silicon in tetramethylammonium hydroxide (TMAH) has been carried out. In TMAH baths,, we find that biased (100) silicon etch rates increase 21% over OCP etch rates. For TMAH baths seasoned with silicon, biased silicon etch rates increase to 63% over those at OCP. Electrochemical etching eliminates the growth of hillocks on etching surfaces regardless of etchant pH, [TMAH] or silicon loading, resulting in highly smooth etching surfaces. Potentiostatic and galvanic etching yield similar etch rates and surface consistency.
Graduation date: 2003

碩士
聖約翰科技大學
電子工程系碩士班
101
ABSTRACT Single-point-lateral the anode electrochemical etching silicon porous silicon used herein, is different from the general use of copper backplane power supply to produce porous silicon. The experiment use a diameter of 0.01mm silver wire, bend side length of 35mm x 35mm square-shaped pull out, then in the positive, and the use of heat-resistant tape close to 40mm x 40mm silicon wafer polishing the surface of the central silicon devices in the etching tank, and then 0.1mm silver wire vertically at the top of the specimen being connected to the negative electrode as cathode ray electrochemical etching of porous silicon, by changing the cathode line and the distance of the silicon etching solution ratio and etching time to conduct research and analysis. Then, respectively, using the UV light and light excitation ray photoelectron spectroscopy (PL) observed porous silicon optical excitation light intensity with the band, SEM, FE-SEM and OM to observe the porous siliconthe electrical characteristics of the reaction surface and cross section of micro-structure, and then use the IV measurement instrument to measure light porous silicon structure. Experimental results displayed, the closer the central cathode line region of porous silicon, etching phenomenon more for significantly but also the more intense, and will be formed an obvious porous silicon circle, contrary away from the cathode lines the more far from the region, etching phenomenon the more weakly, the resultant structure is also largenot identical. Electrical characteristic measurements show that the relative resistance of the porous silicon circle is less than the outer circle, is formed inside the circle the polishing phenomenon, the resistance value of the highest; hope this thesis can provide porous Si as the direction of the optical sensor and a reference. Keywords: lateral etching, electrochemical etching, porous silicon, single point photoluminescence

碩士
國立中興大學
光電工程研究所
106
In this thesis, the porous silicon was formed by electrochemical etching of P-type (100) wafer in different hydrofluoric electrolyte concentration at the constant current density of 5 mA/cm2 for different etching time. We observed the SEM image reveals that PSi pores like dendritic structure, and the structure will be formed according to the shorter diffusion length of the diffusion limiting model. Besides, the thickness of PSi will become 4.5 μm after 30 min of etching. Raman spectrum of PSi was found the FWHM was wider than that of the single crystal silicon, so it could be speculate the reason of FWHM became wider due to the increase of porosity. The Porous silicon which was prepared was divided into two parts for research, first part of the samples was demonstrated with ZnO on different cycles by atomic layer deposition technique (ALD), and the deposition of ZnO 100 cycles (21.10 nm) produces the largest blue shift up to 34 nm; and the second part was used a non-toxic InP/ZnS quantum dots were dropped on the PSi surface, and the red shift phenomenon was apparently observed. It was found that the higher QD concentration did not increase the strength. The ZnO 100 cycles were first deposition on QD/PSi surface, and then AZO (15 nm) is plated on top to form n+-type. PL spectrum had shown that the deposited ZnO was much stronger than QD/PSi. Especially, CHF=12.5% with etching time 10 min showed double peaks of 606 nm and 690 nm. It can be inferred that the broad spectrum is caused by the intrinsic defects of ZnO, porous silicon and QD. The device of the first part with metal coated was analyzed by I-V curve and EL spectrum, but it was found that the metal contact did not form an ohmic contact, which resulted in higher resistance contact, so the AZO/ZnO/QD/PSi structure was used to improve this problem. Log(I)-V characteristics of ZnO/PSi structure can know ideality factor. A ideality factor of 5.75 at etching time 30 min. The longer of etching time, the larger surface roughness and series resistance of the device, it caused the ideality factor larger than the ideal diode. In the EL spectrum, the etching time of ZnO (50, 100 cycles)/PSi structure was 25 and 30 min, and the peak near 450 nm was obvious. It could be inferred that this peak was the blue-green band excited by the ZnO intrinsic defects.

碩士
國立臺灣師範大學
機電科技研究所
96
This research developed photo-assisted electrochemical etching (PAECE) system with low-cost light source for fabricating high-density silicon wafer through-holes array. This process is described as followed: high-density through-holes array in silicon is etched by photo-assisted electrochemical etching under various parameters, such as illumination, surfactants, and concentrations, then to improve the through-hole etching fabrication to obtain through-holes array with high etching rate and smooth etching sidewall. The developed technology will be promising for applications of integrated probe array and wafer-level package in the further. Its advantages are described as followed: low-cost system and fabrication, manufacture, high yield, and suitable for semiconductor process. Using PAECE technology to fabricate wafer through-holes array, we can get the structures with high etching rate and smooth etching sidewall through silicon substrate with thickness of 500 um when the etching time reached 16.7 hours. The smallest width of through-hole by PAECE is 21 um, and the highest aspect ratio is 17.7. The related experimental parameters are described as followed: illumination is 18000-32000 lux, chose surfactants are 1 wt.% DC-1, 1 wt.% MA, 2.5 wt.% H2O2 and 1 wt.% Alcohol. The black micro holes array fabricated by PAECE 2 hr with 1 wt.% MA has ultra-low reflectivity 0.43%, and reflectivity of through-holes array also has equal values about 0.4-05%. Results of this research proved that PAECE technology had been able to partially replace the dry etching technology. It has advantages for applications of integrated probe array and interconnection of wafer-level package. After PAECE fabrication, the black micro holes array will be applied to antireflective structure of solar cell to improve the efficiency obviously.

碩士
國立中央大學
機械工程學系
107
In this study, we perform cryogenic laser-assisted electrochemical etching on heavily doped p-type silicon wafer, and observe the influence of low temperature. Since the carrier mobility of silicon wafer is about 120(cm2/V•s) at room temperature, 108(cm2/V•s) at -20℃, the mobility reduces while temperature decrease. Therefore, the etching experiments are performing at room temperature(25℃),-20℃(dry ice), and using scanning electron microscope(SEM), transmission electron microscope(TEM) and photoluminescence(PL) to observe the structure and characteristics of porous silicon. We find that the lower the temperature, the smaller the nanocrystals grain, and the phenomenon of blue-shift is shown in the result of PL.

博士
國立中央大學
機械工程研究所
96
The technique of electrochemical impedance spectroscopy (EIS) diagnosis has been used to investigate the electrochemical kinetics in the systems of (1) direct methanol fuel cell (DMFC) and (2) photo-electrochemical etching on silicon. The results and contributions of this work were summarized as follows. 1. EIS was carried out to monitor the performance of DMFC under a variety of current densities. Based on analysis of the EIS data that depend upon the performing conditions, an innovative model including the qualitative sketch and its quantitative description relying on postulated equivalent circuit (EQC) was established to delineate the reaction mechanism of DMFC on the membrane electrode assembly (MEA). This model provides a satisfactory diagnosis in the performance of DMFC in terms of the EQC sets. One EQC sets comprises elements such as the internal resistance (Rs) at the highest frequency, the high-frequency impedance (Rif /Cif) that is a parallel combination of the interfacial resistance (Rif) and interfacial capacitance (Cif) resultant from the interfaces in the cell, the medium-frequency impedance (Rrxn /Crxn) that is a parallel combination of the resistance (Rrxn) and capacitance (Crxn) resultant from electrochemical reactions, and the low-frequency impedance (LCO /RCO) that is a parallel combination of the resistance (RCO) and inductance (LCO) resultant from the adsorption and relaxation of CO. This postulated model provides a useful tool to diagnose the degradation mechanism for a cell subject to a test of accelerating degradation. Through the diagnosing and the evidences supported by the examinations through instruments such as the electron probe microanalyzer (EPMA), transmission electron microscope (TEM) and X-ray photoelectron spectroscope (XPS), the degradation is major attributed to (a) the increase of Rif and Rrxn resultant from catalytic degradation that may arise from a series of processes including the dissolution of Ru from the anodic catalyst Pt-Ru, the migration of Ru ions to be reduced on the membrane nearby the cathode. The Ru-dissolution leads to a decrease of catalytic activity on the anode that could be confirmed by the technique of CO stripping in company with the observation through EPMA and XPS. The particles reduced on the membrane nearby the cathode were verified by the examination through TEM and EPMA. (b) The increase of internal resistance (Rs) is ascribed to the loss of sulfonic-acid group from the graded membrane near the anode. Membrane degradation possibly arisen from the heat accumulation in a severely acidic environment near the anode derived from cell reactions. The loss of sulfonic acid group was verified by EPMA and XPS analyses. 2. The photo-electrochemical etching on Si (100) surface reveals different SEM morphologies depending on whether or not the HF solution contains ethanol. Finer smooth pores (around 4 μm in diameter) were formed in the presence of ethanol but larger rough pores (around 8 μm in diameter) formed in 2 M HF solution alone during silicon etched at 0.250 V (vs. SCE) under 50W-illumination for 3 h. The characteristic potentials and current such as transition potential (Etrans), half-wave potential (Ep/2), and limiting current density (jlimit), resulted from dc anodic polarization, were the major parameters used in EIS to diagnose the etching system. There appears an extra low-frequency inductive loop in the Nyquist plot for the etching system in the presence of ethanol. This loop is attributed to relaxation of the adsorption of ethanol in the pores. The contact angle between the etching solution and the silicon decreases with increasing the ethanol concentration. Accordingly, ethanol plays a wetting role in the etching process thus forming fine smooth pores.

博士
國立中央大學
機械工程研究所
93
The aim of this work was to build a thermodynamic energy band diagram for the system of n-type Si (100)/HF that is in dynamic equilibrium at the interface. The concept of the diagram was based on the shift of energy levels such as Fermi energy (EF), conduction band energy (Ec), and valence band energy (Ev) before and after the contact of silicon with HF solutions. Through measurements of the open circuit potential (OCP) and flatband voltage (Vfb), the energy band diagram for the Si/HF system was established. This diagram was useful in estimation of the activation energy for the photo-electrochemical etching system. The kinetic study demonstrated that the etching rate of the silicon (1) increases with an increase of illumination power; (2) increases to a maximum with HF from 0.5 to 2.0 M then decreases with further increase of the HF concentration; (3) accelerates in the presence of 5-10 M EtOH to form smooth macropores but decelerates and caues severe side-etching on the pore walls with the concentration of EtOH reaching 15 M. Based on the energy band diagram established and the electrochemical kinetic data measured, the author was in an attempt to make clear the mechanism for the photo-electrochemical reaction of the n-Si/HF system.

碩士
國立海洋大學
電子工程學系
83
In the experiments of etching under illumination,we firstly discovered that passivative potential of n or p type Si under illumination would shift to more cathodic or anodic direction respectively.To get optimized potential of selective illuminating etching and succeed in fabricating n type mesa with flat surface and sharp edge surrounded with p type isolation area,we have studied such parameters related with the shift of pp as temperature and concentration of KOH solution as well as power and wavelength of ligh source . As a result, etching under illumination in KOH is anisotropic ,and this reveals the possibilities of fabrication of three dimensional micromachines by the technology. The other subject to be studied is that we use electrochemical etch-stop techonology to get cantilever beam silicon thin film as a resonant structure ,and then sputter ZnO on the SiO2/Al/Si to fabricate acoustic sensors / transmitters(actuators). We have performed measurements on the piezoelectric device in one-port configuration,i.e., with both detection and transmission of acoustic signals,and also have tested the sensitivity of self- fabricated acoustic sensors by reciprocity principle.Because the acoustic sensors are fabricated on Si substrate,so they are compatible with IC processes of sensor systems and they have larger potential in research.

碩士
國立清華大學
材料科學工程學系
103
This thesis aims to develop several processes of electrochemical etched porous silicon for either energy storage devices or demonstrating its potential for silicon-based solar cells. Silicon, the second most abundant material on earth, has been utilized in a wide range of regimes including batteries, semiconductor industry, and solar cells due to its low cost and well-developed technology. Porous silicon (PSi), as a functional material either due to its intrinsic property or the porous structure, has been presented to be fabricated by electrochemical etching, which is a facile and cost-effective method for producing porous silicon in a large scale. For lithium-ion batteries, the high theoretical capacity makes it a suitable candidate for anode material. However, the volume expansion during lithiation/delithiation limits its cycling performance. In this thesis, SiNPs produced from PSi has been exploited to lithium-ion batteries with excellent capacitance, columbic efficiency, and cycling retention. Porous silicon is further exploited as electrodes of supercapacitors after few-layer of graphene coating, showing competitive specific capacitance and stable cycling retention. In the last part of the results, porous silicon films and quasi-monocrystalline silicon films possessing extraordinary flexibility are exfoliated from silicon wafers, showing good potential for reusing the silicon wafers to reduce cost of Si-based solar cells.

碩士
國立清華大學
材料科學工程學系
106
Silicon, the second most abundant material on earth, has been applied on wide range regimes such as, batteries, semiconductor, and solar cell. Because of its low cost and well-developed technology, silicon can be used in many application. Porous silicon (PSi), due to its intrinsic property or the porous structure, has been presented to be fabricated by electrochemical etching. For the lithium-ion batteries(LIB), the high theoretical capacity for Si is 3597 mAhg-1. However, the volume expansion during lithiation/delithiation limits its cycling performance. The PSi has been exploited to lithium-ion batteries with excellent capacitance, columbic efficiency, and cycling retention. Recently, PSi has drawn considerable attention for sensor applications because of Its luminescence properties, large surface area, and compatibility with silicon-based technologies. Chemical functionalization of the large surface areas, which can be generated in PSi, show the potential for developing a variety of gas sensors. For this work, the PSi layer apply on NO sensor, and use the UV light to improve the performance. It can easy tell the different concentration of the NO from 0.25 ppm to 5 ppm, and the limit of detection (LOD) is 0.35 ppb.

碩士
國立中央大學
材料科學與工程研究所
103
In recent years, researches on porous silicon and related applications are widely applied to semiconductor processing, solar cells, drug testing and food testing. Thus, the research value of porous silicon is widely acknowledged. In 2012, when Laboratory of Nanoclub was conducting electrochemical etching for the P-type silicon wafer, this research team accidentally discovered that the etching rate would decrease if the He-Ne laser (633nm) was synchronously shined on the surface of the wafer. The study extends the existing research results, aiming to further research the influences of laser energy parameters and laser beam wavelengths on electrochemical etching through the use of the 830nm IR laser. This technology can integrate the MEMS processing techniques such as exposure, development and lithography in the future and can replace the etching process of semiconductors.

碩士
國立海洋大學
電機工程學系
86
In this thesis, we first studied the electrochemical silicon etch process with a pn junction. When proceeding the electrochemical etch technique, we found an interesting phenomenon that can used to evaluate the width of depletion region. Since electrochemical wet etch is the most common type of etch stop under external control, we can measure the width of the depletion region in situ by this method. We compared the experimental result to the theoretic calculated depletion width and tried to find out the relationship between them. With more experiments, we can find the calibrated factor to adjust the deviation of experiments. Moreover, we used electrochemical etch stop technique to form a membrane structure, and then used the best depositing and annealing conditions in our laboratory to deposit LNO, PZT, and Au onto the membrane to be our ultrasonic micromotor. We compared the characteristics of a commercial bulk PZT ultrasonic motor and our device. At the ultrasonic motor operating frequency, the vibrational intensity of our sample is lower than that of the commercial ultrasonic motor. Measuring the electric resonant frequency of our sample, it is at 5.94 MHz. It is inapplicable for ultrasonic motor operation. Therefore, how to adjust the electric and mechanic resonant frequencies to be our needing is the main subject in the future.