Using a modular approach, a vascularized tissue construct is created by embedding functional cells within submillimeter-sized collagen cylinders (modules) while the outside surfaces are seeded with endothelial cells (EC). The void spaces created by randomly packing modules into a container form EC-lined perfusion channels. Upon implantation, the tissues are remodeled by and integrated into the host and experience, to some degree, immune and inflammatory responses. This work utilized microfluidic techniques to study and model the tissue remodeling in vitro in the absence of the host response. When the construct’s tortuous perfusion channels were reproduced in poly(dimethylsiloxane) microfluidic devices and lined with EC, perfusion at higher flow rates reduced EC activation and maintained the desired quiescent EC phenotype. When applying these results to collagen constructs, higher flow rates were not achievable due to the weak mechanical properties of collagen. To increase the collagen’s mechanical strength, a semi-synthetic collagen/poloxamine-methacrylate hydrogel was examined but due to its heterogeneous surface composition, there was inadequate EC attachment and the material was deemed unsuitable for this application. Proceeding with lower flow rates, tissues assembled within microfluidic perfusion chambers from EC-seeded collagen modules showed that over the course of 24 hours, perfusion did not significantly increase activation but instead increased KLF2 expression, a transcription factor involved in the establishment of EC quiescence, and disrupted VE-cadherin bonds between adjacent EC. However, after 1 week of perfusion, the majority of EC were lost. To ameliorate this loss, mesenchymal stromal cells (MSC) were embedded within the modules in order to take advantage of their anti-apoptotic and immunomodulation effects. The MSC temporarily mitigated the loss of the EC but did not prevent it. They did, however, take on a phenotype similar to smooth muscle cells and migrated towards the EC. Perhaps this indicates that the combination of EC, MSC and perfusion drives the creation and assembly of pseudo vessels. Together, the microfluidic techniques used in this study to assemble and perfuse modular tissues revealed new insights into the remodeling process and exposed critical issues surrounding the adaptation of the EC to the combination of perfusion, remodeling and changing flow fields.

Using a modular approach, a vascularized tissue construct is created by embedding functional cells within submillimeter-sized collagen cylinders (modules) while the outside surfaces are seeded with endothelial cells (EC). The void spaces created by randomly packing modules into a container form EC-lined perfusion channels. Upon implantation, the tissues are remodeled by and integrated into the host and experience, to some degree, immune and inflammatory responses. This work utilized microfluidic techniques to study and model the tissue remodeling in vitro in the absence of the host response. When the construct’s tortuous perfusion channels were reproduced in poly(dimethylsiloxane) microfluidic devices and lined with EC, perfusion at higher flow rates reduced EC activation and maintained the desired quiescent EC phenotype. When applying these results to collagen constructs, higher flow rates were not achievable due to the weak mechanical properties of collagen. To increase the collagen’s mechanical strength, a semi-synthetic collagen/poloxamine-methacrylate hydrogel was examined but due to its heterogeneous surface composition, there was inadequate EC attachment and the material was deemed unsuitable for this application. Proceeding with lower flow rates, tissues assembled within microfluidic perfusion chambers from EC-seeded collagen modules showed that over the course of 24 hours, perfusion did not significantly increase activation but instead increased KLF2 expression, a transcription factor involved in the establishment of EC quiescence, and disrupted VE-cadherin bonds between adjacent EC. However, after 1 week of perfusion, the majority of EC were lost. To ameliorate this loss, mesenchymal stromal cells (MSC) were embedded within the modules in order to take advantage of their anti-apoptotic and immunomodulation effects. The MSC temporarily mitigated the loss of the EC but did not prevent it. They did, however, take on a phenotype similar to smooth muscle cells and migrated towards the EC. Perhaps this indicates that the combination of EC, MSC and perfusion drives the creation and assembly of pseudo vessels. Together, the microfluidic techniques used in this study to assemble and perfuse modular tissues revealed new insights into the remodeling process and exposed critical issues surrounding the adaptation of the EC to the combination of perfusion, remodeling and changing flow fields.

One the most daunting technical challenges in the realization of biosensors is functionalizing transducing surfaces for the detection of biomolecules. Functionalization is defined as the formation of a bio-compatible interface on the transducing surfaces of bio-chemical sensors for immobilizing and subsequent sensing of biomolecules. The kinetics of functionalization reactions is a particularly important issue, since conventional functionalization protocols are associated with lengthy process times, from hours to days. The objective of this thesis is the improvement of the functionalization protocols and their kinetics for biosensing applications. This objective is realized via modeling and experimental verification of novel functionalization techniques in microfluidic environments. The improved functionalization protocols using microfluidic environments enable in-situ functionalization, which reduces the processing times and the amount of reagents consumed, compared to conventional methods. The functionalization is performed using self-assembled monolayers (SAMs) of thiols. The thiols are organic compounds with a sulphur group that assists in the chemisorption of the thiol to the surface of metals like gold. The two reactions in the functionalization process examined in this thesis are the SAM formation and the SAM/probe molecule conjugation. SAM/probe molecule conjugation is the chemical treatment of the SAM followed by the binding of the probe molecule to the SAM. In general, the probe molecule is selective in binding with a given biomolecule, called the target molecule. Within this thesis, the probe molecule is an antibody and the target molecule is an antigen. The kinetics of the reaction between the probe (antibody) and the target biomolecule (antigen) is also studied. The reaction between an antigen and its antibody is called the immunoreaction. The biosensing technique that utilizes the immunoreaction is immunoassay. A numerical model is constructed using the finite element method (FEM), and is used to study the kinetics of the functionalization reactions. The aim of the kinetic studies is to achieve both minimal process times and reagents consumption. The impact of several important parameters on the kinetics of the reactions is investigated, and the trends observed are explained using kinetic descriptive dimensionless numbers, such as the Damköhler number and the Peclet number. Careful numerical modeling of the reactions contributes to a number of findings. A considerably faster than conventional SAM formation protocol is predicted. This fast-SAM protocol is capable of reducing the process times from the conventional 24-hours to 15 minutes. The numerical simulations also predict that conventional conjugation protocols result in the overexposure of the SAM and the probe molecule to the conjugation reagents. This overexposure consequently lowers conjugation efficiencies. The immunoreaction kinetics of a 70 kilo-Dalton heat shock protein (HSP70) with its antibody in a hypothetical microchannel is also investigated through the FEM simulations. Optimal reaction conditions are determined, including the flow velocity and the surface concentration of the immobilized probes (antibodies). Based on the numerical results and a series of experimental studies, the fast-SAM protocol application is successfully confirmed. Moreover, the optimum reagent concentration for a given one- hour conjugation process time is determined. This functionalization protocol is successfully applied to immobilize the HSP70 antibody on gold surfaces. The use of the fast-SAM protocol and the predicted optimum conjugation conditions result in binding of the HSP70 antibody on gold, with the same or superior immobilization quality, compared to the conventional protocols. Upon implementation of a 70 μm.s^(-1) flow velocity, the reaction is observed to complete in around 30-35 minutes, which is close to the numerically predicted 30 minutes and 16 seconds. This immunoreaction time is considerably less than conventional 4-12 hour processes. The modified in-situ functionalization techniques achieved here are promising for substantially reducing the preparation times and improving the performance of biosensors, in general, and immunoassays, in particular.

Two types of rotary gate microvalves are developed for flow modulation in a microfluidic system that operates at high flow rate and/or uses particulate flow. This research work encompasses design, microfabrication, and experimental evaluation of these microvalves in three distinct areas compliant micromechanism, microfluidics, and electromagnetic actuation. The microvalve consists of a suspended gate that rotates in the plane of the chip to regulate flow through the orifices. The gate is suspended by a novel fully-compliant in-plane rotary bistable micromechanism (IPRBM) that advantageously constraints the gate in all other degrees of freedom. Multiple inlet/outlet orifices provide flexibility of operating the microvalve in three different flow/port configurations. The suspended gate is made of a soft magnetic material and is electromagnetically actuated like a rotor in a variable-reluctance stepper motor. Therefore, an external electromagnetic (EM) actuation at the integrated set of posts (stator) causes the gate mass to switch from its default angular position to a second angular position. The microvalve chip is fabricated by electroplating a soft magnetic material, Permalloy (Ni80Fe20) in a sacrificial photoresist mold on a Silicon substrate. The inlet/outlet orifices are then etched into the Silicon substrate from the back-side using deep-reactive ion etch process. Finally, the gate structure is released by stripping the PR and seed layers. This research work presents the realization of a new microvalve design that is distinct from traditional diaphragm-type microvalves. The test results are encouraging and show the potential of these microvalves in effectively modulating flow in microfluidic systems that may not require a tight seal. The microvalve uses a novel in-plane rotary bistable micromechanism that may have other applications such as optical shutters, micro-locks, and passive check valves.

Le stockage géologique du CO2 dans les aquifères salins représente une stratégie prometteuse pour la réduction des émissions de CO2 anthropiques. Ce type de stockage requiert des connaissances fondamentales afin d'évaluer les scénarios d'injection, d'estimer la capacité des réservoirs et les risques de fuite. C'est dans ce contexte que des outils microfluidiques haute pression/haute température ont été développés afin d'étudier différents mécanismes liés aux technologies de stockage de CO2. Les laboratoires géologiques sur puce (GLoCs) permettent de mener des expériences à des conditions de pression et de température typiques des réservoirs (25 < T < 50°C, 50 < p < 10 MPa) et d'en mimer des propriétés géologiques.Ce manuscrit présente dans un premier temps les stratégies de fabrication des GLoCs et l'accès à leurs caractéristiques (porosité et perméabilité). La détection du CO2 directement au sein de GLoCs grâce à l'implémentation de fibres optiques par spectroscopie infrarouge est ensuite présentée, ainsi que la mise en œuvre de la technique de laminographie X menées à l'ESRF pour le suivi de dissolution de carbonates dans des microcanaux. Le manuscrit explicite ensuite les investigations, menées avec des GLoCs concernant les différents mécanismes de piégeage du CO2 à l'échelle du pore. La visualisation directe et le traitement d'image ont permis de suivre l'évolution des phases de CO2 et des phases aqueuses au sein du GLoC et les mécanismes de déplacement et de saturation. Enfin, les travaux en cours sont exposés, comme les expériences de drainage avec des saumures réactives ou la formation d'hydrates au sein de milieux poreux. / CO2 geological storage in deep saline aquifers represents a mediation solution for reducing the anthropogenic CO2 emissions. Consequently, this kind of storage requires adequate scientific knowledge to evaluate injection scenarios, estimate reservoir capacity and assess leakage risks. In this context, we have developed and used high pressure/high temperature microfluidic tools to investigate the different mechanisms associated with CO2 geological storage in deep saline aquifers. The silicon-Pyrex 2D porous networks (Geological Labs On Chips) can replicate the reservoir p,T conditions (25 < T < 50°C, 50 < p < 10 MPa), geological and topological properties. This thesis manuscript first highlights the strategies developed during this work to fabricate the GLoCs and to access to global characteristics of our porous media such as porosity and permeability, which are later compared to numerical modelling results. The carbon dioxide detection in GLoCs mimicking p,T conditions of geological reservoirs by using the direct integration of optical fiber for IR spectroscopy is presented. I then detail the strategies for following the dissolution of carbonates in GLoCs with X-rays laminography experiments.Then, the manuscript focuses on the use of GLoCs to investigate each CO2 trapping mechanism at the pore scale. The direct optical visualization and image processing allow us to follow the evolution of the injected CO2/aqueous phase within the reservoir, including displacement mechanisms and pore saturation levels.Eventually, I present the ongoing works such as experiments with reactive brines and hydrates formations in porous media.

The continuous flow separation of magnetic particles from a mixture of particles could improve the performance of magnetic bead based assays but the formation of agglomerates limit the separation efficiency. Bead agglomerates are formed as a result of magnetic binding forces while the hydrodynamic fluid environment strongly influences their movement. The ability to predict the interaction between nearby beads will help to determine a threshold separation distance which will be recommended for use when obtaining measurement within a magnetic bead assay for a specified time interval. The introductory part of this thesis explored the development of a two dimensional numerical model in Matlab which predicts the trajectory pattern as well as magnetic induced velocities between a pair of super-paramagnetic beads suspended in water within a uniform field. The movement of a bead pair interacting due to both magnetic and hydrodynamic forces within a magnetophoretic system was recorded using an optical system; the beads' movements were compared with the simulated trajectories and gave a good agreement. The model was used to predict the shortest agglomeration time for a given separation distance which is of practical benefit to users of bead based assays. The concluding part of this thesis expanded the simulation into a three dimensional model to predict the interactions among three super-paramagnetic beads within a magnetophoretic system. In order to determine the height of the magnetic beads, a Huygens-Fresnel model was implemented in Matlab which was compared with off-focused diffracted images of the beads viewed under an optical system. A good comparison was obtained by comparing the simulated three-dimensional trajectories with experimental data.

L’apparition des objets connectés intelligents dont nous sommes les témoins depuis quelques années a généré un besoin croissant d’antennes à bas coût et énergétiquement sobres. La capacité d’effectuer à la volée une mise en forme du faisceau où sa reconfiguration est une propriété particulièrement intéressante, qui pourrait permettre à l’objet intelligent d’effectuer des tâches telles que la surveillance de zone par exemple ou bien d’optimiser son bilan de liaison en ne visant qu’une seule direction de l’espace. Cela pourrait également mener à un accroissement de l’autonomie de l’objet, via une diminution de sa consommation énergétique, voir à le rendre totalement indépendant s’il devient suffisamment économe pour envisager son alimentation via des systèmes de récupération d’énergie. C’est dans ce contexte que nous proposons ici une nouvelle architecture d’antenne reconfigurable, capable d’un balayage de faisceau sur 360° degrés et basée sur l’utilisation de métal liquide au sein d’un système d’actionnement microfluidique. Dans le premier chapitre, nous ferons une rapide présentation des deux principales technologies de balayage de faisceau utilisées aujourd’hui avant d’étudier les diverses techniques de déplacement de métal liquide utilisées et documentées dans la littérature. L’objectif de ce travail est de sélectionner la technique la plus adaptée à nos besoins. Dans le second chapitre, nous proposerons les deux designs d’antennes envisagés pour notre système, basés sur l’architecture Yagi-Uda. Nous discuterons des avantages et inconvénients de chacun afin d’en sélectionner un qui sera examiné plus en avant dans le chapitre suivant. Dans le troisième chapitre, nous étudierons, à l’aide de simulations électromagnétiques, les performances du design d’antenne sélectionné dans le but de justifier notre choix. Cette étude se concentrera sur l’implémentation graduelle de la complexité du design retenu, en partant d’un système très théorique pour aboutir à une émulation très proche de ce que pourrait être un prototype final. Finalement, dans le quatrième et dernier chapitre nous considérerons deux preuves de concept du système complet ainsi que leurs différentes techniques de fabrications. Étant donné le fait que chaque preuve de concept se concentre soit sur l’aspect RF ou fluidique du système, nous étudierons aussi leurs performances respectives. Nous détaillerons également le développement de certains procédés de fabrication spécifiques utilisés pour réaliser les briques de base, en particulier les objets micro-fluidiques. Ce chapitre nous permet de conclure positivement cette étude de la faisabilité du concept proposé et développé dans ce travail. / The advent of autonomous connected smart objects we are witnessing since a few years has generated a growing need for low cost and energetically sober reconfigurable antennas. The ability to perform on the fly beam shaping and re-configuration is a particularly interesting property which would allow the smart object to perform task such as area surveillance for example and to optimize its link budget by targeting a specific direction of space. This could also allow the increase of the object’s autonomy, through a diminution of its power consumption, or even to render it fully autonomous if it becomes sober enough to envision the use of energy harvesting systems. It is in this context that we propose here a new reconfigurable antenna architecture, capable of 360° beam steering, based on the use of liquid metal within a microfluidic actuation system.In the first chapter, we will do a quick presentation of today’s two main beam steering technics used for antennas before studying the various used and documented technics of liquid metal displacement used in the literature for RF applications. The objective is to single out the better suited one to our requirements.In the second chapter, we will propose the two antenna designs envisioned for our system, based on the Yagi-Uda architecture. We will discuss the advantages and drawbacks of each in order to select one design which will be more closely investigated on the following chapter.In the third chapter, we will study, with the help of electromagnetic simulations, the performances of this selected antenna design in order to justify our choice. This study will focus on the gradual complexity implementation of the chosen design, from a very theoretical system to one very close to what a final prototype would be. Finally, in the fourth and last chapter we will consider two proofs of concept of the complete system and their various fabrications technics. Given that each proof of concept focus either on the RF or the fluidic aspect of the system, we will investigate their performances. We will also detail the development of some of the specific fabrication processes used for the basic building blocks, especially for the fluidic objects. This chapter allow us to conclude positively this study on the feasibility of this concept which was proposed and developed in this work.

Dans cette thèse, nous nous sommes intéressés à la manipulation par forces acoustiques de particules et de fluide à petite échelle. Nous avons construit pour cela un système où des ondes acoustiques de surface sont générées sur un substrat piézo-électrique de LiNbO3 à partir d’électrodes interdigitées, puis émises dans une cavité microfluidique, à une fréquence de l’ordre de 37 MHz soient des longueurs d’onde d'environ 100 µm.Dans le cas où deux ondes stationnaires sont émises perpendiculairement et à la même fréquence, nous montrons théoriquement et expérimentalement la présence d’un terme d’interférence qui, selon le déphasage temporel entre les deux ondes, va modifier la localisation des ventres et nœuds de pression dans la cavité, mais aussi donner lieu à des tourbillons dont l’axe de rotation est perpendiculaire au substrat.Nous montrons théoriquement que ces tourbillons proviennent de la forme particulière des écoulements redressés en paroi et, en injectant des microparticules, nous avons déterminé des vitesses angulaire de plusieurs rad/s. Leur disposition spatiale suit une périodicité d'une demi-longueur d'onde, et leur sens de rotation est alternée entre tourbillons voisins horaires et anti-horaires. Que cela soit avec des globules rouges ou des particules de latex, nous avons identifié une dynamique complexe, avec la formation d’agrégats au centre des vortex sous l’effet des forces de radiations et une répartition en différents niveaux par effet de lévitation acoustique dans l’épaisseur de la cavité, en accord avec l'analyse.Dans le cas où des particules d’une dizaine de micromètres sont utilisées, nous observons, outre l’arrangement des objets dans les nœuds de pression, une rotation individuelle de chaque objet, à des vitesses angulaires plus élevées. Nous interprétons ces observations comme la première mise en évidence d’un couple d’origine acoustique sur des microparticules et cellules biologiques à partir d’ondes acoustiques de surface, constituant l’analogue à petite échelle des effets de couples acoustiques décrits par Busse et Wang en 1981.La thèse propose une description détaillée des différentes montages électriques et microfluidiques, avec les différentes étapes conduisant à un laboratoire sur puce permettant la génération tant de forces que de couples acoustiques, mais aussi la manière de qualifier électriquement et optiquement ses performances. / The focus of this PhD thesis was on particles and fluid handling through acoustic forces, at a very small scale. For this purpose, we built a micro-system based on surface acoustic waves emitted from interdigitated electrodes on a lithium niobate piezoelectric substrate. Those waves then leak into a fluid contained in a microfluidic cavity, at a frequency of 37 MHz, leading to 100 µm wavelengths.If two stationnary waves are emitted perpendicularly and at the same frequency, we theoretically and experimentally show evidence of interferences that can, depending on the time phase shift between them, nto only alter the positions of pressure nodes and antinodes in the acoustic cavity, but also give birth to acoustic vortices which axis is normal to the substrate surface.We theoretically show that those vortices come from the special behaviour of acoustic streaming due to a moving surface. Then, while injecting microparticles in the cavity, we measure angular velocities of a few rad/s. Those vortices spatial disposition follows a half-wavelength period, and their rotation alternates between neighbours: clockwise or anticlockwise. We identify a complex dynamic concerning their 3D structure, since small particles tend to aggregate in vertical columns in the center of the vortex because of radiation forces, with a vertical modulation in the height of the cavity, in good agreement with theoretical predictions.When 10 µm large particles are used instead, we observe individual rotations, even for spherical objects, with higher rotation velocities. We believe those observations to be the first evidence of an acoustic net torque exerted on micro-objects such as biological cells or beads stemming from surface acoustic waves, thus a small scale equivalent of acoustic torques described by Busse and Wang in 1981.This manuscript develops a detailed description of both electric and microfuidic devices, giving the successive steps leading to a lab on chip designed to generate acoustic forces and torques at once, and also the method for qualifying and quantifying electrically and optically its performances.

Cette thèse traite de la fonctionnalisation de surface des résonateurs plasmonique à base de semi-conducteur III-V en utilisant de l’acide phosphonique pour la spectroscopie vibrationnelle exaltée permettant d'identifier des quantités infimes de molécules. Le premier chapitre décrit les fondements théoriques de la spectroscopie vibrationnelle exaltée. En comparant les propriétés plasmoniques du semi-conducteur dégénéré InAs(Sb):Si et des métaux, ici l’or et le gallium, on trouve que l’InAs(Sb):Si est particulièrement adapté à la spectroscopie infrarouge exaltée (SEIRA) et que le gallium est adapté à la spectroscopie Raman exaltée (SERS). Les deux matériaux plasmoniques alternatifs surpassent théoriquement l'or dans leurs gammes spectrales respectives. Néanmoins, l'or et son inertie chimique restent intéressants pour permettre la spectroscopie vibrationnelle exaltée dans différents environnements chimiques.Dans le deuxième chapitre on démontre que l’InAs(Sb):Si est chimiquement stable dans l'eau, contrairement au GaSb. Une structure en couches composites de GaSb/InAsSb:Si a été utilisée pour montrer que la déplétion de l'antimoine et l'incorporation d'oxygène à l'interface GaSb-eau transforment, en un peu moins de 14 h, 50 nm de GaSb cristallin en un oxyde de gallium. Cet oxyde de gallium a un indice de réfraction moyen-IR de l'ordre de n=1,6 ce qui est environ la moitié de la valeur de l'indice de réfraction du GaSb dans le moyen-IR.Dans le troisième chapitre, on démontre que cette modification de l'indice de réfraction lors de l'oxydation peut être exploitée pour décaler la résonance plasmonique localisée des réseaux InAsSb:Si sur des substrats GaSb dans la plage de 5 µm à 20 µm par formation d’un piédestal.Dans le chapitre 4 est présenté le contrôle de la liaison chimique des molécules organiques avec la fine couche d'oxyde natif à la surface du semi-conducteur III-V. L’attachement de ces molécules sur l’oxyde de surface ouvre la voie à des applications bio-photoniques utilisant des semi-conducteurs améliorés par des résonateurs plasmoniques.Dans le chapitre 5 est décrit deux stratégies différentes pour combiner des résonateurs plasmoniques à base de III-V avec des circuits micro-fluidiques. Ces résultats démontrent que des applications lab-on-the-chip basées sur des semi-conducteurs III-V sont possibles.Enfin, la possibilité d'intégrer des nanoparticules de Gallium plasmoniques sur des semi-conducteurs III-V pour combiner les méthodes SEIRA et SERS est présentée au chapitre 6. / This thesis deals with the surface functionalization of nanostructured plasmonic III-V semiconductors for surface-enhanced vibrational spectroscopy relevant to identify minute amounts of analyte molecules.The first chapter outlines the theoretical foundations of surface-enhanced vibrational spectroscopy based on plasmonics. Comparing the plasmonic properties of the degenerate semiconductor InAs(Sb):Si and of metals, here gold and gallium, it is found that the degenerate semiconductor is especially suited for surface-enhanced infrared (SEIRA) spectroscopy and that gallium with its plasmonic potential in the UV-VIS range is apt for surface-enhanced Raman spectroscopy (SERS). Both alternative plasmonic materials theoretically outperform gold in their respective spectral ranges. Nevertheless, gold and its chemical inertness remain interesting for enabling plasmonic enhanced vibrational spectroscopy in different chemical environments. The influence of aqueous environments on the material properties of III-V semiconductors is addressed in the second and in the third chapter. It is found that InAs(Sb):Si is chemical stable in water, but GaSb is not. A GaSb/InAsSb:Si compound layer structure was used to demonstrate that the depletion of antimony and the incorporation of oxygen at the GaSb-water interface transform 50 nm of crystalline GaSb to a gallium oxide in less than 14 hours. The gallium oxide has a mid-IR refractive index in the order of n=1.6 and thus less than half of the value of the mid-IR refractive index of GaSb. This change in refractive index upon oxidation can be exploited to blue-shift the localized plasmonic resonance of InAsSb:Si gratings on GaSb-substrates in the range from 5 µm to 20 µm by pedestal formation.In Chapter 4, the controlled chemical bonding of organic molecules to the approximately 3 nm thin native oxide layer of III-V semiconductor surfaces by phosphonic acid chemistry is presented. This paves the way for plasmonic enhanced all-semiconductor mid-IR biophotonic applications. In chapter 5, two different, but equally successful strategies to combine III-V based plasmonic resonators with microfluidic circuits are described. These results demonstrate that lab-on-the-chip applications based on III-V semiconductors are possible. Finally, the possibility to integrate plasmonic Gallium nanoparticles onto the III-V material platform for a potential combination of SEIRA and SERS applications is presented in chapter 6.

Les deux systèmes de création d’une contrainte dans les films polymériques ont été développés, chacun répondant à un gradient de gonflement du polymère dans la direction normale au film. Ce gonflement peut être provoqué soit par la présence d’un gradient de densité de réticulation dans la direction normale à la surface (films de poly(4-vinylpyridine) réticulés par UV ou dans les films de chitosan réticulés thermiquement et ioniquement ; ou soit par une pénétration asymétrique de vapeur de solvant dans le film (ici le polydiméthylsiloxane oxydé en surface). Un troisième système polymérique auto-enroulant a également été réalisé par la création d’une contrainte permanente au sein du film de polydiméthylsiloxane, grâce à l’extraction sélective d’un additif non-réticulé, l’huile de silicone. Un modèle théorique du processus d’auto-enroulement, basé sur la théorie linéaire d’élasticité a ainsi pu être proposé. / Two systems of stress creation in the polymer films were developed, each based on the swelling gradient in the direction normal to the film. This swelling may be caused either by the presence of a crosslinking density gradient in the direction normal to the surface (poly (4-vinylpyridine film) crosslinked by UV or in the thermally or ionically crosslinked chitosan films; or by asymmetric penetration of solvent vapor in the film (here polydimethylsiloxane surface-oxidized). A third self-rolling polymeric system has also been realized by the creation of a permanent strain in the polydimethylsiloxane film by selective extraction of a non-cross-linked additive, silicone oil. A theoretical model of self-rolling process based on the linear theory of elasticity has been proposed.