Semi-insulating gallium arsenide (SI-GaAs) has gained great interest in recent years due to its wide application in optoelectronic devices and high-speed integrated circuits. An important feature of SI-GaAs is the high density of deep-level defect states, which control the electrical properties of the substrate by compensating the shallow defects. Over the years, deep-level transient spectroscopy (DLTS) and its variations have been the most effective tools employed for the characterization of deep-level defects. However, most of these techniques require a contact probe and tend to be quite restrictive in their applications’ scope. In this thesis deep-level photo-thermal spectroscopy (DLPTS), an all-optical rate-window-based technique, is presented as a novel noncontact technique for the characterization of deep-level defects in SI-GaAs. The signal-generation mechanism for DLPTS is the super-bandgap excitation of carriers, and the sub-bandgap detection of the defect’s thermal-emission process. Combined with the rate-window detection utilizing lock-in amplifiers, DLPTS measurements are performed in three different modalities: temperature-scan, pulse-rate scan, and time-scan. This work demonstrates that each mode provides unique information about the defect configuration, and, in combination, the modes offer a powerful tool for the study of defect properties and optoelectronic processes in SI-GaAs. A hierarchical carrier-emission theory is proposed to explain the thermal broadening (nonexponentiality) in photo-thermal spectra. The model is studied comparatively with the Gaussian distribution of activation energies, and their similarities demonstrate an ergodic equivalence of random energy distribution and the constrained hierarchical emission process. In addition, a rate-window gated photo-carrier radiometry (PCR) technique is developed. The original diffusion-based PCR theory is modified to reflect the signal domination by trap emission and capture rates in the absence of diffusion. Defect luminescence is collected and analyzed using photo-thermal temperature spectra and resonant detection combined with frequency scans. The study results in the identification of five radiative defect states and the defect-photoluminescence quantum efficiency.

Semi-insulating gallium arsenide (SI-GaAs) has gained great interest in recent years due to its wide application in optoelectronic devices and high-speed integrated circuits. An important feature of SI-GaAs is the high density of deep-level defect states, which control the electrical properties of the substrate by compensating the shallow defects. Over the years, deep-level transient spectroscopy (DLTS) and its variations have been the most effective tools employed for the characterization of deep-level defects. However, most of these techniques require a contact probe and tend to be quite restrictive in their applications’ scope. In this thesis deep-level photo-thermal spectroscopy (DLPTS), an all-optical rate-window-based technique, is presented as a novel noncontact technique for the characterization of deep-level defects in SI-GaAs. The signal-generation mechanism for DLPTS is the super-bandgap excitation of carriers, and the sub-bandgap detection of the defect’s thermal-emission process. Combined with the rate-window detection utilizing lock-in amplifiers, DLPTS measurements are performed in three different modalities: temperature-scan, pulse-rate scan, and time-scan. This work demonstrates that each mode provides unique information about the defect configuration, and, in combination, the modes offer a powerful tool for the study of defect properties and optoelectronic processes in SI-GaAs. A hierarchical carrier-emission theory is proposed to explain the thermal broadening (nonexponentiality) in photo-thermal spectra. The model is studied comparatively with the Gaussian distribution of activation energies, and their similarities demonstrate an ergodic equivalence of random energy distribution and the constrained hierarchical emission process. In addition, a rate-window gated photo-carrier radiometry (PCR) technique is developed. The original diffusion-based PCR theory is modified to reflect the signal domination by trap emission and capture rates in the absence of diffusion. Defect luminescence is collected and analyzed using photo-thermal temperature spectra and resonant detection combined with frequency scans. The study results in the identification of five radiative defect states and the defect-photoluminescence quantum efficiency.

Precise manipulation of micro objects became great interest in engineering and science with the advancements in microengineering and microfabrication. In this thesis, a magnetically levitated microgripper is presented for microhandling tasks. The use of magnetic levitation for positioning reveals the problems associated with modeling of complex surface forces and the use of jointed parts or wires. The power required for the levitation of the microgripper is generated by an external drive unit that makes further minimization of the gripper possible. The gripper is made of a biocompatible material and can be activated remotely. These key features make the microgripper a great candidate for manipulation of micro components and biomanipulation. In order to achieve magnetic levitation of microrobots, the magnetic field generated by the magnetic levitation setup is simulated. The magnetic flux density in the air gap region is improved by the integration of permanent magnets and an additional electromagnet to the magnetic loop assembly. The levitation performance is evaluated with millimeter size permanent magnets. An eddy current damping method is implemented and the levitation accuracy is doubled by reducing the positioning error to 20.3 µm. For a MEMS-compatible microrobot design, the electrodeposition of Co-Ni-Mn-P magnetic thin films is demonstrated. Magnetic films are deposited on silicon substrate to form the magnetic portion of the microrobot. The electrodeposited films are extensively characterized. The relationship between the deposition parameters and structural properties is discussed leading to an understanding of the effect of deposition parameters on the magnetic properties. It is shown that both in-plane and out-of-plane magnetized films can be obtained using electrodeposition with slightly differentiated deposition parameters. The levitation of the electrodeposited magnetic samples shows a great promise toward the fabrication of levitating MEMS devices. The end-effector tool of the levitating microrobot is selected as a microgripper that can achieve various manipulation operations such as pulling, pushing, tapping, grasping and repositioning. The microgripper is designed based on a bent-beam actuation technique. The motion of the gripper fingers is achieved by thermal expansion through laser heat absorption. This technique provided non-contact actuation for the levitating microgripper. The analytical model of the displacement of the bent-beam actuator is developed. Different designs of microgripper are fabricated and thoroughly characterized experimentally and numerically. The two microgripper designs that lead to the maximum gripper deflection are adapted for the levitating microrobot. The experimental results show that the levitating microrobot can be positioned in a volume of 3 x 3 x 2 cm^3. The positioning error is measured as 34.3 µm and 13.2 µm when electrodeposited magnets and commercial permanent magnets are used, respectively. The gripper fingers are successfully operated on-the-fly by aligning a visible wavelength laser beam on the gripper. Micromanipulation of 100 µm diameter electrical wire, 125 µm diameter optical fiber and 1 mm diameter cable strip is demonstrated. The microgripper is also positioned in a closed chamber without sacrificing the positioning accuracy.

Precise manipulation of micro objects became great interest in engineering and science with the advancements in microengineering and microfabrication. In this thesis, a magnetically levitated microgripper is presented for microhandling tasks. The use of magnetic levitation for positioning reveals the problems associated with modeling of complex surface forces and the use of jointed parts or wires. The power required for the levitation of the microgripper is generated by an external drive unit that makes further minimization of the gripper possible. The gripper is made of a biocompatible material and can be activated remotely. These key features make the microgripper a great candidate for manipulation of micro components and biomanipulation. In order to achieve magnetic levitation of microrobots, the magnetic field generated by the magnetic levitation setup is simulated. The magnetic flux density in the air gap region is improved by the integration of permanent magnets and an additional electromagnet to the magnetic loop assembly. The levitation performance is evaluated with millimeter size permanent magnets. An eddy current damping method is implemented and the levitation accuracy is doubled by reducing the positioning error to 20.3 µm. For a MEMS-compatible microrobot design, the electrodeposition of Co-Ni-Mn-P magnetic thin films is demonstrated. Magnetic films are deposited on silicon substrate to form the magnetic portion of the microrobot. The electrodeposited films are extensively characterized. The relationship between the deposition parameters and structural properties is discussed leading to an understanding of the effect of deposition parameters on the magnetic properties. It is shown that both in-plane and out-of-plane magnetized films can be obtained using electrodeposition with slightly differentiated deposition parameters. The levitation of the electrodeposited magnetic samples shows a great promise toward the fabrication of levitating MEMS devices. The end-effector tool of the levitating microrobot is selected as a microgripper that can achieve various manipulation operations such as pulling, pushing, tapping, grasping and repositioning. The microgripper is designed based on a bent-beam actuation technique. The motion of the gripper fingers is achieved by thermal expansion through laser heat absorption. This technique provided non-contact actuation for the levitating microgripper. The analytical model of the displacement of the bent-beam actuator is developed. Different designs of microgripper are fabricated and thoroughly characterized experimentally and numerically. The two microgripper designs that lead to the maximum gripper deflection are adapted for the levitating microrobot. The experimental results show that the levitating microrobot can be positioned in a volume of 3 x 3 x 2 cm^3. The positioning error is measured as 34.3 µm and 13.2 µm when electrodeposited magnets and commercial permanent magnets are used, respectively. The gripper fingers are successfully operated on-the-fly by aligning a visible wavelength laser beam on the gripper. Micromanipulation of 100 µm diameter electrical wire, 125 µm diameter optical fiber and 1 mm diameter cable strip is demonstrated. The microgripper is also positioned in a closed chamber without sacrificing the positioning accuracy.

La présente thèse porte sur l’étude de l’auto-organisation induite par laser de nanoparticules (NPs) d’argent dans des couches minces de TiO2. Cette technique laser a été récemment mise au point pour fournir des solutions d'impression originales pour des applications comme l'imagerie polarimétrique, la sécurité ou le design. Ici, nous étudions expérimentalement et théoriquement les différents mécanismes impliqués dans la formation des NPs pour un meilleur contrôle de leur morphologie et de leur organisation. Dans le cas de lasers continus, nos résultats expérimentaux prouvent que la vitesse à laquelle le faisceau laser balaye la surface de l’échantillon est un paramètre crucial pour le contrôle de la taille finale des NPs, et de l’élévation de température induite dans la couche par le laser. La taille des NPs d’argent diminue aux basses vitesses d’écriture, alors que leur croissance est activée par effet thermique et apparait uniquement au-dessus d’un seuil de vitesse. La taille finale des NPs varie dans ce cas de façon non-monotone avec la vitesse de balayage alors que la hausse de température ne cesse de croître. Afin d'expliquer ces phénomènes contre-intuitifs, plusieurs modèles numériques ont été développés qui permettent de simuler la variation de la taille des NPs sous insolation laser dans le domaine spectral du visible. Les résultats théoriques concordent bien avec les observations expérimentales et révèlent que l’origine de ces phénomènes résulte d'interactions complexes entre les différents processus physico-chimiques impliqués dans la formation des NPs. Finalement, cette thèse démontre que l’auto-organisation des NPs à l’échelle sub-longueur d’onde peut être aussi réalisée avec des impulsions laser femto-secondes (fs), et une croissance anisotrope des NPs a été également observée pour une gamme particulière de paramètres laser. La comparaison avec les structures de NPs produites par lasers continus suggère que les impulsions laser fs offrent une meilleure maîtrise des effets thermiques et de la taille des NPs. L'usage de laser fs offre ainsi une nouvelle approche pour transférer ces technologies sur des supports souples comme les papiers ou les plastiques / This doctoral thesis focuses on the study of laser-induced self-organization of Ag nanoparticles (NPs) in TiO2 thin layers. This laser technique was recently developed to provide original printing solutions for applications like active color displays, security, or polarization imaging. Here, we investigate experimentally and theoretically the different mechanisms involved in the NPs formation for a better control of their morphology and organization. In the case of continuous-wave (cw) laser, our experimental results prove that the speed at which the laser scans the sample surface is a crucial parameter to control the NP size and the laser-induced temperature rise. Ag NPs shrink at low scan speed, whereas growth governed by a thermal effect only occurs above a speed threshold. Above this threshold, the size of grown Ag NPs changes in a non-monotonous way with scan speed, whereas laser-induced temperature rise increases with speed. In order to explain these counter-intuitive behaviors, several numerical models have been developed to simulate changes in the size distribution of Ag NPs induced by visible light. Simulation results are in good agreement with experimental observations and reveal the origin of the observed phenomena by collective interactions of various physico-chemical processes involved in the variation of NPs size. Finally, this thesis demonstrates that self-organization of NPs at sub-wavelength scale on large areas can also be performed with femtosecond (fs) laser pulses. Oriented anisotropic growth of NPs has also been observed using a proper selection of laser irradiation parameters. Compared with the cw laser induced structures, fs pulses offer better control of the thermal effect and NP size, which opens the way to transfer these technologies on plastic or paper substrates

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / The objectives of this thesis were: To estimate the relative maturity group of old and modern soybean cultivars used in Southern Brazil; To quantify the branches contribution in the leaf area index; To characterize the development and the overlap period of vegetative and reproductive phases, growth in height and the emission of nodes after the beginning of flowering in determinate and indeterminate plant type cultivars in different sowing dates and soybean regions in Rio Grande do Sul and; To identify key biophysical and management factors governing variation in soybean yield potential in Southern Brazil. Were conducted 47 experiments using 23 soybean cultivars in five growing seasons (2010/2011, 2011/2012, 2012/2013, 2013/2014 and 2014/2015) at 12 sites in Rio Grande do Sul. The relative maturity group these cultivars that had not yet been classified ranged from 4.8 to 7.3. Most cultivars showed a reduction on total leaf area index, maximum, of main stem and of branches, in the duration of the reproductive phase, of the total development cycle with delay on sowing date regardless of the maturity group and type of growth. The indeterminate type cultivars had a higher overlap period of vegetative and reproductive phases, growing in height and emission nodes after the beginning of flowering bigger than determinate type cultivars. The potential yield ranged from 2.8 up to 6.0 Mg ha-1, depending on the amount and distribution of water during the growing season and the photothermal coefficient during reproductive stages of the soybean. / Os objetivos desta tese foram: Estimar o grupo de maturidade relativa de cultivares antigas e modernas de soja; Quantificar a contribuição das ramificações no índice de área foliar; Caracterizar o desenvolvimento e a duração do período de sobreposição das fases vegetativa e reprodutiva, o crescimento em estatura e a emissão de nós após o início do florescimento em cultivares modernas de soja com diferentes grupos de maturidade relativa, tipos de crescimento, semeadas em épocas e regiões produtoras de soja no Rio Grande do Sul e; Identificar as variáveis meteorológicas e de manejo que governam o potencial de rendimento em soja no sul do Brasil. Foram conduzidos 47 experimentos utilizando 23 cultivares de soja, durante cinco estações de crescimento (2010/2011, 2011/2012, 2012/2013, 2013/2014 e 2014/2015) em 12 locais no Rio Grande do Sul. O grupo de maturidade relativa das cultivares que ainda não tinham sido classificadas variou de 4.8 a 7.3. Há redução dos valores de índice de área foliar máximo, da haste principal e das ramificações, da duração da fase reprodutiva e do ciclo total de desenvolvimento, com o atraso da época de semeadura, independentemente do grupo de maturidade relativa e tipo de crescimento. As cultivares com tipo de crescimento indeterminado apresentam maior período de sobreposição das fases vegetativa e reprodutiva, crescimento em estatura e emissão de nós após o início do florescimento do que as cultivares determinadas. O potencial rendimento de soja variou de 2,8 até 6,0 Mg ha-1, dependendo da quantidade e distribuição da água durante a estação de crescimento e do coeficiente fototérmico durante os estágios reprodutivos da soja.

The path towards mitigation of anthropogenic greenhouse gas emissions lies in the transition from conventional to sustainable energy resources. The Hydrogen Economy, a cyclic economy based on hydrogen as a fuel, is suggested as a tool in the necessary energy transition. Photocatalysis makes use of sunlight to promote thermodynamically non-favoured reactions such as water splitting, allowing for sustainable hydrogen production. Harvesting thermal energy along with photonic energy is an interesting concept to decrease the activation energy of water splitting (i.e. ΔG = + 237.2 kJ∙mol−1). This work aims to confront this hypothesis in a gas phase photo-thermal reactor designed specifically for this study. The photocatalyst chosen is graphitic carbon nitride (g-C3N4), an organic semiconductor possessing a narrow band gap (i.e. 2.7 eV) as well as a band structure which theoretically permits water splitting. The photocatalytic performance of Pt/g-C3N4 for hydrogen evolution was tuned by altering its synthetic temperature. Electron paramagnetic resonance was used to gain insight on the evolution of the photocatalyst activity with synthesis temperature. Then, gold nanoparticles were deposited on g-C3N4 surface. Localized surface plasmon resonance properties of gold nanoparticles are reported in the literature to be influenced by temperature. Therefore Au/g-C3N4 appeared as a promising candidate for photo-thermal water splitting. X-ray spectroscopy unveiled interesting observations on the gold oxidation state. Moreover, under specific reduction conditions, gold nanoparticles with a wide variety of shapes characterized by sharp edges were formed. Finally, the development of the photo-thermal reactor is presented. The design process and the implementation of this innovative reactor are discussed. The reactor was successfully utilized to probe photoreactions. Then, the highly energy-demanding photocatalytic water splitting was proven not to be activated by temperature in the photo-thermal apparatus.

La conduction thermique devient moins efficace à mesure que la taille des struc-tures diminuent en desous du micron, car la diffusion de surface des phononsdevient prédominante et limite plus efficacement les phonons que la diffusionphonon-phonon Umklapp. Des études récentes ont indiqué que les phonon po-laritons de surface (SPhPs), qui sont les ondes électromagnétiques évanescentesgénérées par l’hybridation des phonons optiques et des photons et se propageantà la surface d’une surface diélectrique polaire, pourraient servir de nouveauxvecteurs de chaleur pour améliorer les performances thermiques dans des dis-positifs micro- et nano-métriques. Nous étudions l’état des SPhPs existantdans un film submicronique diélectrique dans une large gamme de fréquences.Le calcul de la conductivité thermique des SPhPs basé sur l’équation de trans-port de Boltzmann (BTE) montre que le flux de chaleur transporté par lesSPhPs est supérieur à celui des phonons. Nous effectuons également une mesurede réflectance thermique dans le domaine temporel (TDTR) de films submi-croniques deSiNet démontrons que la conductivité thermique due aux SPhPsà haute température augmente lorsque l’épaisseur du film dimine. Les résultatsprésentés dans cette thèse ont des applications potentielles dans le domaine dutransfert de chaleur, de la gestion thermique, du rayonnement en champ proche et de la polaritoniques. / Thermal conduction becomes less efficient as structures scale down into submicron sizes since phonon-boundary scattering becomes predominant and impede phonons more efficiently than Umklapp scattering. Recent studies indicated that the surface phonon polaritons (SPhPs), which are the evanescent electromagnetic waves generated by the hybridation of the optical phonons and the photons and propagating at the surface of a polar dielectric material surface, potentially serve as novel heat carriers to enhance the thermal performance in micro- and nanoscale devices. We study the condition of SPhPs existing in a dielectric submicron film with a broad frequency range. The calculaton of SPhPs thermal conductivity based on Boltzmann transport equation (BTE) demonstrates that the heat flux carried by SPhPs exceeds the one carried by phonons. We also conduct a time-domain-thermal-reflectance (TDTR) measurement of $SiN$ submicron films and demonstrate that the thermal conductivity due to the SPhPs at high temperatures increases by decreasing the film thickness. The results presented in this thesis have potential applications in the field of heat transfer, thermal management, near-field radiation and polaritonics.

Chickpea (Cicer arietinum L.) was recently introduced to the Canadian prairies, a region which has a short growing season in which crop maturation often occurs under cool and wet conditions. To improve the yield of chickpea, crop duration must closely match the available growing season. The objectives of this study were to: i) examine the days to flowering of diverse chickpea accessions grown in either long or short-days; ii) examine the days to flowering of selected chickpea accessions grown in a range of thermal regimes combined with either long or short days and to examine the interaction between photoperiod and day and night temperatures on crop duration; iii) determine the timing and duration of the photoperiod-sensitive phase in selected chickpea accessions, and vi) determine the genetic basis of the association between flowering time and reaction to ascochyta blight in chickpea. A wide variation was observed in chickpea accessions for their response to flowering under long (16/8 hours day /night) and short days (10/14 hours day/night). Earlier flowering was observed under long photoperiod regimes compared with the short photoperiod regimes. Variability was detected among chickpea accessions for their flowering responses when different temperatures were combined with different photoperiods. Earlier flowering was observed under long days (16/8 hours day/night) coupled with high to moderate temperature regimes (24/16 ºC and 20/12 ºC, day and night respectively) compared to short-days (10/14 hours day and night) and moderate to low temperature regimes (20/12 ºC and 16/8 ºC day and night, respectively). Those chickpea accessions such as ICC 6821 and ICCV 96029 which originated from the lower latitudes of Ethiopia and India, respectively, flowered earlier compared to accessions such as CDC Corinne and CDC Frontier which originated from the higher latitudes and cooler temperate environments of western Canada. Photoperiod sensitivity phases were detected in chickpea accessions adapted to the cold environments of western Canada, whereas no photoperiod sensitivity phase was identified in the extra-early flowering cultivar ICCV 96029. The duration of the photoperiod sensitive phase in the chickpea accessions was longer under short days compared to long days. Field and growth chamber evaluation of a chickpea RIL population (CP-RIL-1) revealed the presence of variability among the lines and the two parents for their days to flowering and level of resistance to ascochyta blight. Broad sense heritability across different site-years for days to flower 0.45 to 0.78, plant height 0.48 to 0.78, ascochyta blight resistance 0.14 to 0.68, days to maturity 0.26, photoperiod sensitivity 0.83 and nodes number of first flowering 0.37 to 0.75 were estimated. Days to flower and photoperiod sensitivity were significantly r = -0.21 to -0.58 (P ≤ 0.05 to 0.001) and -0.28 to -0.41 (P ≤ 0.01 to 0.001), respectively and negatively correlated with ascochyta blight resistance in the CP-RIL-1 population. A genetic linkage map consisting of eight linkage groups was developed using 349 SNP markers. Seven QTLs were identified for days to flowering under growth chamber and field conditions on chromosomes 3, 5, 6 and 8 each and 3 QTLs on chromosome 4. The total phenotypic variation explained by QTLs for days to flowering ranged from 7 to 44%. Two QTLs for days to maturity were identified on chromosomes 3 and 8. Three QTLs, one each on chromosomes 3, 4 and 5 were identified for photoperiod sensitivity. The total phenotypic variation explained by each QTL for photoperiod sensitivity ranged from 7 to 41%. A total of three QTL for node of first flowering, one on chromosomes 3 and 8 each, and two on chromosome 4 were identified. The two QTL on chromosome 4 explained total phenotypic variations of 11 and 32%, respectively. Ten QTLs distributed across all chromosomes, except chromosomes 2 and 5, were identified for ascochyta blight resistance. The phenotypic variability explained by each QTL for ascochyta blight resistance ranged from 7 to 17%. The molecular markers associated with these QTLs have potential for use in chickpea breeding.

Les plasmons-polaritons de surface (PPS) font figure de plateforme polyvalente très promet- teuse pour le guidage des ondes électromagnétiques à l’échelle nanométrique. Dans ce contexte, le contrôle dynamique de la propagation PPS est d’une importance capitale. Le contrôle actif des dispositifs plasmoniques a souvent été réalisé jusqu’à présent par le biais d’un effet thermo-optique (TO). Toutefois dans la majorité des cas considérés, l’effet thermo-optique résulte d’une modification des propriétés d’un matériaux diélectrique en contact avec le métal supportant le mode plasmon. Ainsi, le rôle des propriétés thermo-optiques du métal lui-même a rarement été analysé aux fréquences télécom dans le cadre d’applications plasmoniques. L’objectif principal de cette thèse est donc d’analyser en détail l’impact des propriétés thermo- optiques des métaux sur différents modes PPS aux longueurs d’ondes télécom. En premier lieu, nous considérons la modulation photo-thermique d’un mode plasmon supporté par un film mince d’or se propageant à l’interface "or/air". Nous démontrons tout d’abord la modulation de la propagation des modes PPS induite par la dépendance des pertes ohmiques de l’or à la température du film mince. Le contrôle de la température du film est obtenu par un effet photo-thermique en régime continu modulé. Les mesures expérimentales de la pro- fondeur de modulation de l’intensité des modes PPS combinées à la simulation numérique de la distribution de température le long du film d’or nous permettent de remonter aux coefficients thermo-optiques de l’or aux fréquences télécoms. Dans un second temps, nous considérons le contrôle thermo-optique de modes plasmons dont le confinement spatial (et donc l’indice effectif) est supérieur à ceux des modes de films. Les modes considérés dans cette seconde étude sont connus sous le nom de "polymer- loaded surface plasmon waveguides (PLSPPWs)". Ces modes présentent un confinement latéral induit par l’indice de réfraction du ruban de polymère déposés sur le film métallique et un confinement vertical résultant de leur nature plasmonique. L’excitation photo-thermique de ces guides plasmoniques dans un régime nano-seconde nous permet de mettre en évidence la dynamique thermo-optique du métal aux temps courts (<1ns) et du polymère aux temps plus longs (<1µs). La même démarche appliquée à un micro-résonateur plasmonique en anneau révèle les temps caractéristiques de la dynamique de diffusion de la chaleur dans le polymère à l’échelle de quelques dizaines de nanomètres. Sur la base de ces expériences, nous suggérons un design de dispositifs plasmoniques thermo-optique dont la bande passante est de l’ordre du megahertz, un ordre de grandeur au-dessus des systèmes thermo-optiques traditionnels. Enfin, nous présentons la modulation photo-thermique de la propagation PPS le long de nanofils d’or fabriqués par lithographie électronique supportant des modes plasmons très confinés aux longueurs d’onde télécom. La transmission d’un signal télécom à 10 Gbit/s est tout d’abord démontrée afin d’établir sans ambiguïté la pertinence de tels guides d’ondes miniatures pour la transmission d’informations à très courtes échelles. Enfin, nous mettons en évidence la modulation photo-thermique de la propagation de tels modes. En particulier, nous investiguons l’influence sur la profondeur de modulation de la polarisation du faisceau pompe relativement à l’orientation des nanofils. Cet effet de polarisation s’explique par une absorption exaltée si la polarisation du faisceau pompe est orientée perpendiculairement à l’axe du nanofil. L’exaltation résulte de l’excitation d’un mode plasmon local selon l’axe transverse du nanofil. / Surface plasmon polaritons (SPPs) is the promising versatile platform proposed for guiding electromagnetic waves at nanoscale dimensions. In this context dynamic control of SPPs prop- agation is of paramount importance. Thermo-optical (TO) effect is considered as an efficient technique for performing active control of plasmonic devices. Among the thermo-optical based plasmonic devices demonstrated so far TO coefficient is dominantly provided by a dielectric material on top of the metal sustaining the SPP mode, however, the role of TO properties of the metal has been rarely investigated for plasmonic applications especially at the telecom frequency ranges. Therefore, the aim of this thesis is to investigate in detail the impact of thermo-optical properties of metals onto various SPP modes at telecom wavelengths.First, we report on photo-thermal modulation of thin film SPP mode traveling at gold/air interface excited at telecom wavelengths. We start by investigating the photo-thermally in- duced modulation of SPPs propagation mediated by the temperature dependent ohmic losses in the gold film. Then we extract the thermo-plasmonic coefficient of the SPP mode from the accurately measured SPPs signal depth of modulation by which we could compute the thermo-optical coefficients (TOCs) of gold at telecom wavelength. Lastly, we demonstrate a pulsed photo-thermal excitation of the SPPs in the nanosecond regime.Secondly, we investigate the thermo-optical dynamics of polymer loaded surface plasmon waveguide (PLSPPW) based devices photo-thermally excited in the nanosecond regime. First, we demonstrate thermo-absorption of PLSPPW modes mediated by the temperature-dependent ohmic losses of the metal and the thermally controlled field distribution of the plasmon mode within the metal. Next, we consider the thermo-optical response of a PLSPPW based racetrack shaped resonator coupled to a straight bus waveguide and evaluate the photo-thermal activation through heating and cooling times. We conclude that nanosecond excitation combined to high thermal diffusivity materials opens the way to high speed thermo-optical plasmonic devices.Finally, we report on the photo-thermal modulation of SPPs propagation along litho- graphically fabricated gold nanowires sustaining highly confined plasmonic mode at telecom wavelengths. First, we investigate telecommunication characterization of the nanowires by ap- plying high bit rate signal transmission, 10 Gbit/s, through fiber-to-fiber confocal detection setup. Next, we demonstrate and evaluate the photo-thermal modulation of SPPs propagation along the nanowires where we discuss qualitatively TO effects due to light-induced modula- tions on nanowires and show the impact of the incident beam polarization on the photo-thermal modulation.