No description available.

The applicability of the technique of inductively coupled plasma mass spectrometry to the analysis of geological samples was investigated using a variety of sample introduction techniques including: solution nebulisation; slurry nebulisation; flow injection; electrothermal vaporisation; and laser ablation, Solution sample introduction is limited by the amount of time required to prepare the sample, and the relative intolerance of the technique to high concentrations of sample matrix. The maximum level of dissolved solids for a refractory matrix such as a digested igneous rock was found to be 0.2% w/v. Good accuracy and precision are achievable. Acceptable results can be obtained using slurry nebulisation. However, standardisation is a problem due to the difference in response for aqueous and slurried analytes. Calibration against aqueous standards and the use of an internal standard is therefore precluded. In addition, the preparation of stable slurries is a highly skilled and time consuming task. Flow injection analysis offers the most benefit to the geochemical analyst. Flow injection was found to increase sample throughput and, more important, to improve matrix tolerance by a factor of l0x, thus allowing the direct determination of the platinum group elements and gold in geological samples without pretreatment. Small samples, such as fluid inclusion leachates can also be analysed without dilution and the matrix effects experienced when analysing samples containing high salt concentrations can be reduced by careful control of dispersion. The potential for increased detection limits by electrothermal vaporisation was not proved for geological materials due to the same matrix tolerance problems which limit detection limits in solution work. Laser ablation sampling allows direct analysis of the solid but quantitation requires matrix matched standards or independent variable internal standardisation, limiting the applicability of the technique for bulk screening. The use of laser ablation to analyse trace element concentrations in individual mineral grains has been investigated and partition coefficients for trace elements, including the rare earths in a large zoned pyroxene crystal, were determined.

The Raman Injection Laser is a new type of laser which is based on triply resonant stimulated Raman scattering between quantum confined states within the active region of a Quantum Cascade Laser that serves as an internal optical pump. The Raman Injection Laser is driven electrically and no external laser pump is required. Triple resonance leads to an enhancement of orders of magnitude in the Raman gain, high conversion efficiency and low threshold. We studied this new type of laser and conclude some basic equations. With reasonable experimental parameters, we calculated the laser gain, losses and the output power of the Raman Injection Laser by using Mathematica and FEMLab. Finally we compared the theoretical and experimental results.

The Raman Injection Laser is a new type of laser which is based on triply resonant stimulated Raman scattering between quantum confined states within the active region of a Quantum Cascade Laser that serves as an internal optical pump. The Raman Injection Laser is driven electrically and no external laser pump is required. Triple resonance leads to an enhancement of orders of magnitude in the Raman gain, high conversion efficiency and low threshold. We studied this new type of laser and conclude some basic equations. With reasonable experimental parameters, we calculated the laser gain, losses and the output power of the Raman Injection Laser by using Mathematica and FEMLab. Finally we compared the theoretical and experimental results.

One of two project reports. Part A can be found at: http://hdl.handle.net/11375/17866 / <p> It was observed that the spontaneous carrier lifetimes of GaAlAs injection lasers, as measured by the time delay method, exhibit a strong dependence on the DC bias current flowing through the laser. As the DC bias current is increased, the measured lifetimes decrease, passing through a minimum at bias currents less than 0.1 mA, and then rise again to an intermediate value at high bias currents. It was also found that the impedance of these lasers to fast current pulses drops sharply at the same DC bias currents as the minimum measured lifetime mentioned above. Evidence is presented to suggest that both of these effects are due to an electrical resonance between the stray drive circuit inductance and the DC bias dependent capacitance of the laser's junction under forward bias. </p> <p> In addition, two methods of determining the spontaneous carrier lifetime by other than the time delay method are described. One method involves measurement of the vector impedance of lasers to sinusoidal drive currents at frequencies less than 100 MHz, as a function of DC bias. The second method is based upon the frequency dependence of the phase shift between a sinusoidal drive current and the resulting spontaneous light. </p> / Thesis / Master of Engineering (ME)

La vision à travers les milieux diffusants est rendue difficile à cause des interactions entre la lumière et les particules composant le milieu. Dans le cadre de la vision dans le brouillard, ces trvaux de thèse portent sur l'étude de deux techniques optiques : l'une fondée sur la tomographie par cohérence optique (OCT) couplée à un cristal photo-réactif (PCR) et l'autre reposant sur le principe de la rétro-injection dans un micro-laser. Un modèle analytique fondé sur la théorie de Mie est développé pour prédire la décroissance du rapport signal à bruit dans une scène bruitée par du brouillard. L'exploitation des résultats analytique met en avant la nécessité de sélectionner l'information utile et d'éliminer le bruit lumineux lors des prises de vue pour conserver des images correctement contrastées . Nous démontrons le fort potentiel exploitable des deux techniques d'imagerie étudiées puisqu'elles reposent sur le principe de sélection du signal utile. Le montage OCT-PRC réalise une sélection par cohérence temporelle et le montage de rétro-injection laser une sélection spatiale. Les résultats obtenus en présence d'un milieu perturbateur sont détaillés en explicitant les limitations des dispositifs actuels et les voies d'améliorations à entreprendre.

Le travail décrit dans cette thèse porte sur une nouvelle méthodologie de caractérisation des circuits sécurisés basée sur une rétro-conception matérielle partielle : d’une part afin d’améliorer l’injection de fautes laser, d’autre part afin de détecter la présence de Chevaux de Troie Matériels (CTMs). Notre approche est dite partielle car elle est basée sur une seule couche matérielle du composant et car elle ne vise pas à recréer une description schématique ou fonctionnelle de l’ensemble du circuit.Une méthodologie invasive de rétro-conception partielle bas coût, rapide et efficace est proposée. Elle permet d’obtenir une image globale du circuit où seule l’implémentation des caissons des transistors est visible. La mise en œuvre de cette méthodologie est appliquée sur différents circuits sécurisés. L’image obtenue selon la méthodologie déclinée précédemment est traitée afin de localiser spatialement les portes sensibles, voire critiques en matière de sécurité. Une fois ces portes sensibles identifiées, nous caractérisons l’effet du laser sur différentes parties de ces cellules de bases et nous montrons qu’il est possible de contrôler à l’aide d’injections de fautes laser la valeur contenue dans ces portes. Cette technique est inédite car elle valide le modèle de fautes sur une porte complexe en technologie 90 nm. Pour finir une méthode de détection de CTMs est proposée avec le traitement de l’image issue de la rétro-conception partielle. Nous mettons en évidence l’ajout de portes non répertoriées avec l’application sur un couple de circuits. La méthode permet donc de détecter, à moindre coût, de manière rapide et efficace la présence de CTMs. / The work described in this thesis covers an integrated circuit characterization methodology based on a partial hardware reverse engineering. On one hand in order to improve integrated circuit security characterization, on the other hand in order to detect the presence of Hardware Trojans. Our approach is said partial as it is only based on a single hardware layer of the component and also because it does not aim to recreate a schematic or functional description of the whole circuit. A low cost, fast and efficient reverse engineering methodology is proposed. The latter enables to get a global image of the circuit where only transistor's active regions are visible. It thus allows localizing every standard cell. The implementation of this methodology is applied over different secure devices. The obtained image according to the methodology declined earlier is processed in order to spatially localize sensible standard cells, nay critical in terms of security. Once these cells identified, we characterize the laser effect over different location of these standard cells and we show the possibility with the help of laser fault injection the value they contain. The technique is novel as it validates the fault model over a complex gate in 90nm technology node.Finally, a Hardware Trojan detection method is proposed using the partial reverse engineering output. We highlight the addition of few non listed cells with the application on a couple of circuits. The method implementation therefore permits to detect, without full reverse-engineering (and so cheaply), quickly and efficiently the presence of Hardware Trojans.