Ce travail concerne le développement d’un nouveau type de microscopie optique en champ proche (SNOM) basé sur la mise en œuvre de sondes dite actives qui utilisent le signal de génération de seconde harmonique (SHG) d’un petit nombre de molécules orientées. L’orientation de ces molécules est obtenue par l’application d’un champ électrique statique dans une jonction constituée d’une pointe métallique effilée placée à proximité d’un substrat conducteur et immergée dans une solution de molécules dipolaires non-linéaires. L’excitation laser de ces molécules localement orientées permet d’obtenir une polarisation non-lineaire à fréquence double qui constitue une nanosource de lumière intrinsèquement localisée et pouvant interagir avec le champ proche du substrat. Nous nous sommes intéressés à l’imagerie de nano-objets lithographiés par cette technique de SNOM-SHG. Nous avons pu démontrer la possibilité d’obtenir une résolution de l’ordre de 200 nm, soit une résolution meilleure d’un facteur 2 par rapport à la limite de diffraction.Nous avons ensuite étudié les moyens d’optimiser les performances de ce nouveau type de sondes SNOM-SHG. Une voie consiste à exploiter les propriétés d’antenne optique de pointes métalliques effilées, qui peuvent être le siège d’effets d’exaltation du champ électromagnétique résultant de la singularité géométrique de ces objets (extrémité effilée) ou de l’excitation de résonances plasmons. Afin de pouvoir quantifier ces effets, nous avons entrepris la caractérisation, par luminescence à 2 photons (TPL), de nanofils d’or considérés comme objets de référence pour mimer une pointe. Des fils lithographiés ainsi que des fils issus de chimie colloïdale ont été étudiés de façon à mieux comprendre à la fois l’influence de la forme et de la cristallinité des objets sur les exaltations de champ. Des études simultanées de la géométrie et des propriétés optiques d'un nanofil unique ont été menées au moyen d'un microscope optique inversé associé à une excitation laser et couplé à un microscope à force atomique (AFM) dont la pointe est préalablement réglée pour coïncider avec le spot laser. En balayant l’échantillon, nous pouvons directement confronter l’image topographique de l’objet à la cartographie de points chauds enregistrés à sa surface, le signal de TPL étant directement corrélé à la densité locale d’états électromagnétiques. Nous avons pu montrer que les fils lithographiés et les fils colloïdaux présentaient des facteurs d’exaltation locale de champ différents, la cristallinité des objets pouvant aussi être révélée que via l’analyse spectrale du signal de TPL émis. Enfin, un dernier volet important de mon travail a consisté à faire évoluer le banc expérimental précédemment développé au laboratoire de façon à pouvoir réaliser simultanément des caractérisations de type SNOM-SHG et des caractérisations topographiques. Dans ce but, nous avons travaillé à l’intégration d’une tête AFM diapason sur notre banc de microscopie non-linéaire. Au-delà des aspects électroniques liés à l’optimisation du fonctionnement de ce diapason, le couplage du faisceau laser dans le microscope a également été entièrement reconfiguré. / This work deals with the development of a new kind of scanning near-field optical microscopy (SNOM) based on the realization of so-called active probes taking advantage of the second harmonic generation (SHG) signal coming from a few oriented molecules. The orientation of these molecules is obtained by applying a static electric field in a junction made of a sharp metallic tip placed close to a conductive substrate and immersed in a solution containing dipolar non-linear molecules. A second order nonlinear polarization is obtained from these locally oriented molecules following their excitation with a laser beam finally leading to a nanosource of light intrinsically localized and able to interact with the near-field of the substrate.We have investigated this SNOM-SHG technique to image nano-objects made by e-beam lithography. We were able to demonstrate that a resolution of about 100 nm could be reached, which appears better (of a factor2) than the diffraction limit.We have then been focusing on the way to improve the capabilities of this new type of SNOM-SHG probes. One approach consists in taking advantage of the optical antenna effects that can occur at the end of sharp tips, where the electromagnetic field can be enhanced due to geometrical effects (sharp extremities) or due to the excitation of plasmon resonances. In order to quantify these field enhancements, we have carried out the characterization of gold nanowires using two-photon luminescence (TPL) ; considering these wires as reference objects that can mimic tips. Nanowires made by e-beam lithography and nanowires synthesized by colloidal chemistry have both been studied in order to have a better understanding of the influence of the shape and the crystallinity on the field enhancements. Simultaneous analysis of the geometry and the optical properties of a single nanowire has been carried out using an inverted microscope associated to a laser excitation and coupled to an atomic force microscopy (AFM) which tip is previously aligned with the laser spot. When scanning the sample, we can directly correlate the topographic image of the object to the mapping of the hotspots recorded on its surface, the TPL signal being directly linked to the electromagnetic local density of states. We were able to evidence that both nanowires made by e-beam lithography or synthesized by colloidal chemistry exhibit different field enhancement factors, the crystallinity of the objects being also revealed following the spectral analysis of the emitted TPL signal.Finally, a last important part of my work has dealt with the evolution of the experimental setup previously developed in the laboratory in order to be able to achieve simultaneously SNOM-SHG type and topographic characterizations. We have therefore been working on the integration of an AFM tuning fork head to our nonlinear optical bench. Above the electronic aspects related on the optimization of the tuning fork implementation, the coupling of the laser beam in the microscope has also been reconfigured.

Effective focusing of electromagnetic (EM) energy to nanoscale regions is one of the major challenges in nano-photonics and plasmonics. The strong localization of the optical energy into regions much smaller than allowed by the diffraction limit, also called nanofocusing, offers promising applications in nano-sensor technology, nanofabrication, near-field optics or spectroscopy. One of the most promising solutions to the problem of efficient nanofocusing is related to surface plasmon propagation in metallic structures. Metallic tapered rods, commonly used as probes in near field microscopy and spectroscopy, are of a particular interest. They can provide very strong EM field enhancement at the tip due to surface plasmons (SP’s) propagating towards the tip of the tapered metal rod. A large number of studies have been devoted to the manufacturing process of tapered rods or tapered fibers coated by a metal film. On the other hand, structures such as metallic V-grooves or metal wedges can also provide strong electric field enhancements but manufacturing of these structures is still a challenge. It has been shown, however, that the attainable electric field enhancement at the apex in the V-groove is higher than at the tip of a metal tapered rod when the dissipation level in the metal is strong. Metallic V-grooves also have very promising characteristics as plasmonic waveguides. This thesis will present a thorough theoretical and numerical investigation of nanofocusing during plasmon propagation along a metal tapered rod and into a metallic V-groove. Optimal structural parameters including optimal taper angle, taper length and shape of the taper are determined in order to achieve maximum field enhancement factors at the tip of the nanofocusing structure. An analytical investigation of plasmon nanofocusing by metal tapered rods is carried out by means of the geometric optics approximation (GOA), which is also called adiabatic nanofocusing. However, GOA is applicable only for analysing tapered structures with small taper angles and without considering a terminating tip structure in order to neglect reflections. Rigorous numerical methods are employed for analysing non-adiabatic nanofocusing, by tapered rod and V-grooves with larger taper angles and with a rounded tip. These structures cannot be studied by analytical methods due to the presence of reflected waves from the taper section, the tip and also from (artificial) computational boundaries. A new method is introduced to combine the advantages of GOA and rigorous numerical methods in order to reduce significantly the use of computational resources and yet achieve accurate results for the analysis of large tapered structures, within reasonable calculation time. Detailed comparison between GOA and rigorous numerical methods will be carried out in order to find the critical taper angle of the tapered structures at which GOA is still applicable. It will be demonstrated that optimal taper angles, at which maximum field enhancements occur, coincide with the critical angles, at which GOA is still applicable. It will be shown that the applicability of GOA can be substantially expanded to include structures which could be analysed previously by numerical methods only. The influence of the rounded tip, the taper angle and the role of dissipation onto the plasmon field distribution along the tapered rod and near the tip will be analysed analytically and numerically in detail. It will be demonstrated that electric field enhancement factors of up to ~ 2500 within nanoscale regions are predicted. These are sufficient, for instance, to detect single molecules using surface enhanced Raman spectroscopy (SERS) with the tip of a tapered rod, an approach also known as tip enhanced Raman spectroscopy or TERS. The results obtained in this project will be important for applications for which strong local field enhancement factors are crucial for the performance of devices such as near field microscopes or spectroscopy. The optimal design of nanofocusing structures, at which the delivery of electromagnetic energy to the nanometer region is most efficient, will lead to new applications in near field sensors, near field measuring technology, or generation of nanometer sized energy sources. This includes: applications in tip enhanced Raman spectroscopy (TERS); manipulation of nanoparticles and molecules; efficient coupling of optical energy into and out of plasmonic circuits; second harmonic generation in non-linear optics; or delivery of energy to quantum dots, for instance, for quantum computations.