No description available.

Cette thèse est dédiée à l'extension du Tracé de Rayons Vectoriels Complexes (TRVC) développé au laboratoire et son application à la caractérisation des particules non-sphériques. Dans divers domaines de recherche, tels que la mécanique des fluides et la combustion, nous devons mesurer les caractéristiques des particules. Parmi les différents types de techniques de mesure, la métrologie optique est largement utilisée grâce à sa précision et sa fiabilité. Cependant, la forme des particules est souvent considérée comme sphérique principalement à cause du manque de moyens pour prédire avec précision l'interaction de la lumière avec de grandes particules non-sphériques. TRVC a été développé pour répondre à ces besoins. Il est basé sur le modèle de rayons mais une amélioration radicale a été apportée dans ce nouveau modèle par l'introduction d'une nouvelle propriété dans la notion de rayons lumineux - la courbure de front d'onde. Dans cette thèse, le modèle est appliqué à réexaminer la théorie d'Airy. Il a été montré que TRVC prédit mieux les intensités et les positions des lobes secondaires dans les arcs-en-ciel d'une particule sphérique que la théorie d'Airy. Ensuite, TRVC est appliqué à l'étude des propriétés des arcs-en-ciel formés par les ellipsoïdes allongé et aplati. Il est montré que TRVC peut prédire analytiquement les positions et les intensités de lobes secondaires dans les arcs-en-ciel d’une particule sphéroïdale lorsqu’elle est éclairée par une onde plane dans le plan symétrique. Les pics dus à l'effet focal d'une particule sphéroïdale aplatie ont également été évalués analytiquement en utilisantle facteur de divergence de TRVC. Un système de mesure est aussi mis en place dans le laboratoire pour la diffusion de la lumière par des gouttes pendantes. Il a été montré que le rapport des intensités des deux premiers arcs-en-ciel est sensible à l'ellipticité d'un sphéroïde équivalent. Ceci ouvre un champ d'application potentiel pour caractériser la déformation d'une goutte pendante. / This thesis is dedicated to the extension of Vectorial Complex Ray Model (VCRM)developed in the laboratory and its application in the characterization of large nonspherical particles. In various research domains, such as the uid mechanics and the combustion field, we need to measure the characteristics of the particles. Among di_erent kinds of measurement techniques, the optical metrology is largely employed due to its advantage of being accurate, reliable and non-intrusive. However, in many optical techniques, the shape of the particles is assumed to be spherical. The main reason of this limit is due to the lack of theoretical model to describe with precision the interaction of light with large non-spherical particles. The Vectorial Complex Ray Model has been developed to reply this demand. It is based on the ray model but a radical improvement has been done by introducing a new property the wave front curvatures in the ray model. In this thesis, the model is firstly applied to reexamine the Airy theory. It is shown that even in the case of spherical particle VCRM predicts better the intensity and positions of supernumerary bows of rainbow than the Airy theory. Then VCRM is applied to investigate the properties of the rainbows formed by a spheroidal (prolate or oblate) particle. It is shown that VCRM can predict analytically the positions and the intensity of supernumerary bows and the peaks due to the focal effect when it is illuminated by a plane wave in the symmetrical plane.The theoretical research results have been also applied to the experimental characterization of a pendant drop. The intensity ratio of the two first orders of rainbow is shown sensible to the aspect ratio of the equivalent spheroid and it opens a potential to develop a measurement technique to characterize the deformation of pendant drop.

No description available.

Manufacturing of precise structures in MEMS, semiconductors, optics and other fields requires high standards in manufacturing and quality control. Appropriate surface topography measurement technologies should therefore deliver nm accuracy in the axial dimension under typical industrial conditions. This work shows the characterization of a dispersion-encoded low-coherence interferometer for the purpose of fast and robust surface topography measurements. The key component of the interferometer is an element with known dispersion. This dispersive element delivers a controlled phase variation in relation to the surface height variation which can be detected in the spectral domain. A laboratory setup equipped with a broadband light source (200 - 1100 nm) was established. Experiments have been carried out on a silicon-based standard with height steps of 100 nm under different thermal conditions such as 293.15 K and 303.15 K. Additionally, the stability of the setup was studied over periods of 5 hours (with constant temperature) and 15 hours (with linear increasing temperature). The analyzed data showed that a height measurement of 97.99 ± 4:9nm for 293.15 K and of 101.43 ± 3:3nm for 303.15 K was possible. The time-resolved measurements revealed that the developed setup is highly stable against small thermal uctuations and shows a linear behaviour under increasing thermal load. Calibration data for the mathmatical corrections under different thermal conditions was obtained.

This work introduces a modified low-coherence interferometry approach for nanometer surface-profilometry. The key component of the interferometer is an element with known dispersion which defines the measurement range as well as the resolution. This dispersive element delivers a controlled phase variation which can be detected in the spectral domain and used to reconstruct height differences on a sample. In the chosen setup, both axial resolution and measurement range are tunable by the choice of the dispersive element. The basic working principle was demonstrated by a laboratory setup equipped with a supercontinuum light source (Δλ = 400 ̶ 1700 nm). Initial experiments were carried out to characterize steps of 101 nm on a silicon height standard. The results showed that the system delivers an accuracy of about 11.8 nm. These measurements also served as a calibration for the second set of measurements. The second experiment consisted of the measurement of the bevel of a silicon wafer. The modified low-coherence interferometer could be utilized to reproduce the slope on the edge within the previously estimated accuracy. The main advantage of the proposed measurement approach is the possibility to collect data without the need for mechanically moving parts.

Within this work a scan-free, low-coherence interferometry approach for surface profilometry with nm-precision is presented. The basic setup consist of a Michelson-type interferometer which is powered by a supercontinuum light-source (Δλ = 400 - 1700 nm). The introduction of an element with known dispersion delivers a controlled phase variation which can be detected in the spectral domain and used to reconstruct height differences on a sample. In order to enable scan-free measurements, the interference signal is spectrally decomposed with a grating and imaged onto a two-dimensional detector. One dimension of this detector records spectral, and therefore height information, while the other dimension stores the spatial position of the corresponding height values. In experiments on a height standard, it could be shown that the setup is capable of recording multiple height steps of 101 nm over a range of 500 µm with an accuracy of about 11.5 nm. Further experiments on conductive paths of a micro-electro-mechanical systems (MEMS) pressure sensor demonstrated that the approach is also suitable to precisely characterize nanometer-sized structures on production-relevant components. The main advantage of the proposed measurement approach is the possibility to collect precise height information over a line on a surface without the need for scanning. This feature makes it interesting for a production-accompanying metrology.

Zur vollständigen Beschreibung eines monochromatischen Wellenfeldes ist die Kenntnis über die Amplituden- und Phasenverteilung unabdingbar. Während sich die messtechnische Erfassung der Amplitudenverteilung durch lichtempfindliche Sensoren recht einfach realisieren lässt, gestaltet sich die Bestimmung der Phasenverteilung weitaus schwieriger. Die Phasenverteilung eines optischen Wellenfeldes kann nur über indirekte Verfahren gewonnen werden. Es ergibt sich ein sogenanntes phase retrieval Problem. Zur Lösung dieses Problems bieten sich verschiedene Verfahren aus dem Bereich der berührungslosen und zerstörungsfreien optische Messtechnik an. In dieser Arbeit wird ein deterministisches Verfahren zur Phasenrekonstruktion mit Hilfe Greenscher Funktionen vorgestellt. Die erste Greensche Identität dient als Grundlage zur Entwicklung einer Gleichung, welche in der Lage ist, bei der Rekonstruktion einer Phasenverteilung spezifische Randbedingungen zu berücksichtigen. Dies ermöglicht unter anderem eine genaue Charakterisierung von Phasenobjekten bzw. ihren optischen Eigenschaften, wie beispielsweise der Brechzahlverteilung. Das vorgestellte Verfahren zur Phasenrekonstruktion basiert einerseits auf schnellen Algorithmen, welche die Leistung von parallelen Prozessoren ausnutzen und andererseits auf geschickten experimentellen Aufbauten, mit welchen die notwendigen Eingangsdaten zur Lösung der Gleichung simultan gewonnen werden können. Es ergibt sich damit die Möglichkeit, die Amplituden- und Phasenverteilung eines Wellenfeldes in Echtzeit zu bestimmen und daraus folgend ein Mittel zur quantitativen Bewertungen und Analyse von dynamischen Prozessen sowohl in der Industrie als auch im Bereich der Life Sciences. / In order to describe a monochromatic wave field entirely, knowledge about the amplitude and phase distribution is elementary. While it is easy to measure the amplitude distribution of an optical wave field by the use of photosensitive detectors, the determination of the phase distribution is by far more difficult. Due to the fact, that the phase distribution can not be measured directly, a problem of phase retrieval is presented. This problem may be solved by applying a non-contacting and non-destructive optical metrology technique. In this thesis a deterministic method for phase retrieval based on Green''s functions will be introduced. Green''s first identity serves as a starting point to derive an equation for phase retrieval considering different boundary conditions. Among others, this allows an exact characterization of phase objects, or their optical properties, as for example the refractive index distribution. On the one hand, the presented phase retrieval technique is based on fast algorithms which take advantage of the performance of parallel processors. On the other hand, skilful experimental setups allow the simultaneous acquisition of the input data, which are necessary to solve the phase retrieval equation. It follows that the presented technique is able to determine the amplitude and phase distribution of a wave field in real-time. Hence this technique enables the quantitative evaluation and analysis of dynamic processes in industry as well as in the area of life sciences.

[pt] Objetivando suprimir as instabilidades dinâmicas de ganho em um amplificador à fibra dopada com Érbio (EDFA), uma nova configuração de laser em anel é apresentada e demonstrada. Neste trabalho, analizamos os efeitos da variação do nível de atenuação no laço de re-alimentação sobre a resposta transitória do EDFA. Particularmente, observamos as excursões de ganho experimentadas pelo canal sobrevivente quando sete dentre oito canais da rede são adicionados ou removidos, à exemplo do que ocorre em sistemas WDM reais. Sob esta análise, avaliamos o desempenho do sistema em suprimir as instabilidades dinâmicas de ganho do EDFA. / [en] A new ring laser configuration to eliminate the gain dynamic instabilities of an erbium doped fiber amplifier (EDFA) is proposed and demonstrated. We examine the effect of the attenuation level in the optical feedback path over thetransient response of the EDFA. In particular, we look at the transient gain excursions experienced by surviving channel when seven of eight channels are added or dropped, like in real WDM systems. Using this analysis as a guide, we highlight the robustness of the approach and evaluate its performance to EDFA gain stabilization.